CodeCanyon

Every developer has their own unique style of writing code. Despite companies establishing best practices and drawing up documentation, it can be hard for developers to understand someone else’s codebase. To solve this problem, Dennis Pilarinos built a tool called Unblocked — an AI-powered assistant that answers contextual questions about lines of code. Pilarinos is a seasoned professional at building developer tools. He has worked as a director at Microsoft and Amazon, working on Azure and Amazon Web Services. Later, Pilarinos built a continuous integration platform for remote teams called Buddybuild, which was acquired by Apple in 2018, and he worked at Apple for around two years on the Xcode Cloud platform. “Developers trying to get the information they want is quite painful and time-consuming,” Pilarinos told TechCrunch in a phone interview. “We wanted to use all the data, code, and tribal knowledge of conversations to present developers with easy answers.” Pilarinos added that with more AI programming tools seeping into developers’ workflows, the aforementioned problems are going to get worse. Image Credits:UnblockedUnblocked integrates with development environments and apps like Slack, Jira, Confluence, Google Drive, and Notion. The tool gathers intelligence about a company’s codebase and helps answer questions such as “Where do we define user metrics in our system?” Developers can also use the platform to search for the person who made changes to a particular module and quickly gain insights from them. Unblocked offers admin controls that can be easily adopted by a company’s system admin, and the startup is working on integrating with platforms like Cursor and Lovable to improve code explainability. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW Beyond this, Unblocked is developing tools that actively help developers with projects rather than simply answer questions. One, Autonomous CI Triage, supports developers in testing code through different scenarios. Unblocked counts companies such as Drata, AppDirect, Big Cartel, and TravelPerk as customers. Pilarinos claims that engineers at Drata were able to save one to two hours per week using Unblocked’s platform. Image Credits:UnblockedUnblocked said Tuesday it has raised $20 million in Series A funding from B Capital and Radical Ventures. The round takes the company’s total capital raised to $30 million from investors, including Amplify Partners, First Round Capital, and XYZ Capital. Rob Toews, a partner at Radical Ventures, said that as AI-generated code proliferates, products like Unblocked will become extremely valuable. “There is a thesis that much of the software in the future will be written by AI,” Toews said. “There will be second- and third-order effects of that, creating new challenges. One of them is that people won’t understand when, why, and how [a particular piece of code] was written.” Toews thinks that, just as AI-powered Glean enabled enterprises to make sense of their troves of data, Unblocked will help devs understand their increasingly complicated codebases.

Google on Tuesday announced the launch of Gemini 2.5 Pro Preview (I/O edition), an updated version of its flagship Gemini 2.5 Pro AI model that the company claims tops a number of widely used benchmarks. Gemini 2.5 Pro Preview (I/O edition) is available via the Gemini API and Google’s Vertex AI and AI Studio platforms, and is priced the same as the Gemini 2.5 Pro model it effectively replaces. It’s also in Google’s Gemini chatbot app for the web and for mobile devices. Very excited to share the best coding model we’ve ever built! Today we’re launching Gemini 2.5 Pro Preview 'I/O edition' with massively improved coding capabilities. Ranks no.1 on LMArena in Coding and no.1 on the WebDev Arena Leaderboard. It’s especially good at building… pic.twitter.com/9vRaP6RTTo — Demis Hassabis (@demishassabis) May 6, 2025 The model’s release comes ahead of Google’s annual I/O developer conference (hence the “I/O edition” designation), where Google is expected to unveil a host of models, as well as AI-powered tools and platforms. The company is competing fiercely for mindshare and marketshare in the cutthroat AI race; rivals like OpenAI and xAI are on the cusp of releasing models that are expected to be highly performant. According to Google, Gemini 2.5 Pro Preview (I/O edition) has “significantly” improved capabilities for coding and building interactive web apps. The model is also better at tasks like code transformation — that is, modifying a piece of code to achieve a specific goal — and code editing, the company says. Image Credits:GoogleIn a blog post, Google notes that Gemini 2.5 Pro Preview (I/O edition) leads the WebDev Arena Leaderboard, a benchmark measuring a model’s ability to create aesthetically pleasing and functional web apps. Gemini 2.5 Pro Preview (I/O edition) also has state-of-the-art performance in video understanding, achieving a score of 84.8% on one popular benchmark, VideoMME. “For developers already using Gemini 2.5 Pro, this new version will not only improve coding performance but will also address key developer feedback including reducing errors in function calling and improving function calling trigger rates,” wrote Google in its blog post. “By default, the model has a real taste for aesthetic web development while maintaining its steerability.”

Google’s iOS app is getting a new “Simplify” feature that uses AI to make complex or technical text on the web easier to understand, all without leaving a web page. Simplify features a prompt refinement approach developed by Google Research and uses Gemini to make complicated text more digestible, without losing key details. Google says its testing found that the simplified text allowed users to better understand and retain the original information. By offering a built-in feature like Simplify, Google is likely hoping that people will stay within its ecosystem of tools and services for help with complex topics, rather than turning to popular third-party tools like OpenAI’s ChatGPT. You can use Simplify by selecting any text on a web page you’re visiting in the Google app and then tapping the “Simplify” icon that appears. You’ll then see a simpler version of the text to help you digest the information and allow you to continue reading. Image Credits:GoogleGoogle’s research blog shares an example of what this feature could look like. For example, you may come across a passage that reads: “The complex pathology of this condition involves emphysematous destruction of lung parenchyma, diffuse interstitial fibrosis, changes in the composition of lung immune cells, increased production of immunomodulatory factors, and the prominent remodeling of pulmonary vasculature.” The simplified text would then read: This complex condition involves damage to the lung tissue from emphysema, a disease that damages the air sacs in the lungs, and widespread scarring of the lung tissue, called fibrosis. The immune cells in the lungs change, and the body makes more immunomodulatory factors, substances that control the immune system. The blood vessels in the lungs also change a lot.” Google says the idea behind the feature is to make it easier for people to grasp complex topics they might come across when trying to learn something new on the web. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW “Our goal requires models to paraphrase complex ideas accurately without introducing errors or omitting key details,” Google said in a blog post. “The rewritten text must help the reader understand challenging material without sacrificing the integrity of the original information.” Simplify is rolling out to iOS users this week.

Finom, an Amsterdam-based digital bank for small- and medium-sized businesses, has raised €92.7 million (roughly $105 million) in a growth investment from General Catalyst’s Customer Value Fund, the company tells TechCrunch exclusively. The capital infusion “will be used exclusively and only for growth” and not for operational expenses or product development, Kos Stiskin, Finom’s chairman and co-founder, told TechCrunch. He described it as a nontraditional funding round in which General Catalyst doesn’t take any equity. “[O]ur core operations are generating positive cash flow, and all new investments and funding go directly toward attracting new clients,” Stiskin said. Finom is primarily in the banking business, but this year, the company expanded its offerings beyond digital banking services. In February, Finom unveiled what Stiskin described as an “autonomous AI accounting agent” for entrepreneurs and freelancers in Europe. And in March, the startup expanded into direct lending, which incorporates an AI-powered scoring engine. Finom’s credit offering, available in the Netherlands, will be expanded across Europe by year-end, Stiskin added. Today, Finom counts over 100,000 businesses across Germany, France, Spain, the Netherlands, and Italy as customers, reporting positive unit economics in all markets. Its revenue model is primarily subscription-based. Finom also generates revenue through transaction fees for certain services and offers a competitive cash-back program. The recent expansion into lending also opened a new revenue stream through interest on credit lines. Image Credits:FinomStiskin declined to reveal hard revenue figures, but he told TechCrunch that Finom doubled its annual recurring revenue in 2024 and that the company is “EBITDAM [earnings before interest, taxes, depreciation, amortization, and marketing] profitable.” Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW In an interview, Stiskin described Finom’s closest competitor as Qonto, a Paris-based challenger bank that in January 2022 announced a massive €486 million (~$552 million) Series D funding round. But Stiskin believes that Finom has a “stronger localization strategy and more comprehensive product suite.” Presently, Finom has 505 employees, up 31.5% compared to last year. Last September, the company named Alessandro Camilotti, former head of finance and analytics EU at Klarna, as its CFO. In total, Finom has raised nearly €190 million (roughly $214 million) since its inception in 2020. In February 2024, Finom announced it had raised €50 million (roughly $56 million) in a Series B equity round of funding co-led by General Catalyst and Northzone. The startup has declined to reveal its valuation. According to PitchBook, Finom was valued at $150.7 million post-money in November 2021 after a €30 million (roughly $33.8 million) seed funding round from VCs Target Global, Tal Ventures, and General Catalyst. Zeynep Yavuz, partner at General Catalyst, believes that Finom has “shown strong execution in a market that is still deeply underpenetrated.” She also thinks its modular infrastructure gives the company the ability “to scale efficiently” across geographies, “leveraging shared capabilities while localizing where needed.” “We see Finom’s proprietary anti-money laundering and know-your-customer engine as a standout advantage — not just for compliance, but for customer experience,” Yavuz said.

Particle, the startup behind an AI-powered newsreader that aims to help publishers, not just steal their work, is bringing its product to the web. On Tuesday, the company announced the launch of the new Particle.news website that connects news consumers with headlines and AI summaries from a variety of sources, plus the ability to delve into various categories like Technology, Sports, Entertainment, Politics, Science, Crime, Economics, and Video Games, in addition to browsing the day’s most popular stories on the home page. The company thinks that bringing its product to the web will help to reach more readers, giving them a different way to keep up with the news using AI technology enhancements. Like the existing Particle mobile app, the site offers AI tools designed to help consumers better understand the news. Instead of just summarizing stories into key bullet points for quicker reading, Particle also extracts key quotes and allows users to ask questions about the story via an AI chatbot. These questions and answers from users are available on the new website, but the site doesn’t yet offer the ability to interact with the AI directly. While reading the news on Particle, if you want to learn more about a topic, you can access “entity pages” that detail information about a specific person, product, or organization mentioned in the story. For instance, when you see the word “Trump” or “Knicks” or “Nintendo Switch” highlighted in a headline or news summary, you can click through to a page offering basic information pulled from Wikipedia and links to more stories about the subject. ScreenshotImage Credits:ParticleParticle also highlights the news outlets covering a story by sharing links to their stories directly alongside its AI summaries. In early tests on mobile, the company found that readers were clicking through to the publishers’ sites via these links, leading Particle to begin partnering with specific publishers like Reuters, Fortune, and the AFP to display their links more prominently. On the new website, Particle also displays links to related reporting at the bottom of its AI summaries to keep users clicking through to read more. Plus, when users share a link from Particle’s mobile app, it will connect readers with the dedicated landing page on the website, opening up access to Particle’s content to more readers, including those who don’t have the app installed. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW The addition of AI into the news and journalism market has been controversial at times, particularly when some publishers attempted to outsource reporting to AI bots, leading to much backlash. But Particle’s founders want to find a way to use AI to help readers better understand the news, without stealing traffic from publishers. Particle was founded in 2023 by the former senior director of product management at Twitter, Sara Beykpour, and a former senior engineer at Twitter and Tesla, Marcel Molina. It’s backed by $4.4 million in seed funding and a $10.9 million Series A led by Lightspeed. Particle joins other efforts to leverage AI for news summaries, including TechCrunch’s former parent company, Yahoo, which acquired the Artifact news app from Instagram’s co-founders to revamp its News app with AI-powered features. Bloomberg, Gannett (USA Today), The Wall Street Journal, and others have also been experimenting with AI article summaries. However, readers are likely less forgiving of AI’s mistakes on the news outlets’ websites where they’re directly reporting the news, compared with an independent app devoted to AI summaries and Q&As.

Nearly three years after unveiling the concept version of the flagship Cadillac Celestiq EV, GM has finally released the real thing: a “mid-$300,000” electric vehicle that oozes luxury and a come-gawk-at-me exterior designed to lure in passersby. But is this bespoke EV enough for Cadillac to regain its position as the “Standard of the World”? I spent a day navigating this behemoth around Los Angeles in a bid to find out. Sliding behind the steering wheel of the new flagship fastback is a good start. I am surrounded by buttery leather — my rear end comforted by the cushy seat, which is adjustable in every way and then some. The fit and finish is precise and elevated like a Richard Mille watch and everything I touch has that distinct honed-by-hand aesthetic. And yet in 115 cases, those parts have been 3D printed, including the center of the steering wheel, window switches, parts of the console decor, and even some structural pieces. Cadillac clearly wanted to build an EV with zero compromises. The battery pack layout, however, presented certain design challenges. The modules in the Cadillac Celestiq’s battery pack aren’t uniform. Unlike the Lyriq, which has a flat battery pack, the Celestiq has different configurations of modules in the battery pack that are different heights. The modules under the front seats are about nine inches tall, but the ones under the rear passenger footwell are only six inches in height. The rear seats are brought up with modules that are 12 inches tall. Engineers needed to place some modules under the center console to get to a 111 kWh battery with 303 miles of range. The result is a shallow storage compartment that’s hardly big enough to fit my wallet and glasses. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW Cadillac Celestiq EV: The drive Image Credits:Emme HallThe looks I received while piloting the Celestiq made up for that teeny center console storage. There is nothing on the road that looks this graceful, with its dancing LED light signature in the front, long nose, and sculpted rear end. In a town full of manufactured beauty, the real-deal Celestiq stands out. I start off on the rough streets of Hollywood, where I intentionally aim for broken pavement and manhole covers. My tester is shod with 22-inch wheels — although 23-inch wheels are available — and Michelin Pilot Sport EV tires with very little sidewall. Having this little rubber between the car and the pavement often results in a harsh ride, but the Celestiq sports a compliant air suspension that keeps it all in check. Bigger events like potholes certainly make their presence known, but city driving is quiet and smooth. Make no mistake, this car is a biggun. The Cadillac Celestiq has a larger footprint than a two-door Chevrolet Silverado, made slightly more nimble with rear steering. Sure, it’s a bit hard to find street parking in this bad boy, but this car is more likely to find itself at the valet. Leaving Hollywood, I head to the hills to see just what 655 horsepower and 646 pound-feet of torque can do on a curvy road. Here the Magnetic Ride Control shines, reacting quickly to the road as well as the car’s weight transfer to provide a confident feeling of control. Aided by active roll control, the car feels flat through the turns despite its zaftig proportions. The steering here is a bit numb, but it’s weighted nicely and the brake regen provides a new skill to conquer. I find that at max regen, if I can time my throttle lifts right, I never have to hit the mechanical brakes before heading into a corner. The weight still transfers to the front so I have the grip to turn, but I’m also getting free electrons at the same time. I love it. I crank up the Dolby Atmos sound system pouring out of 38 internal AKG speakers — there are four outside the car that produce cool EV sounds for those unlucky enough to be walking — and head out to the highway. Of course, GM’s Super Cruise advanced driver-assistance system is here, so I take my hands off the wheel and let the car take over for a while in traffic. The Cadillac Celestiq EV has a problem Image Credits:Emme HallAnd therein lies the biggest problem with the Celestiq. I’m not necessarily mad that all GM electric vehicles run Google Built-In over Apple CarPlay and Android Auto. Google Maps is great in this vehicle, accurately predicting my range when I arrive at a location, offering up charging stations and preconditioning the batteries if needed. It’s just that I can’t reliably get the system to hook up with my phone for hands-free texts and phone calls. And it’s not just in the Celestiq. In both the Optiq and the Escalade IQ, the process was persnickety. There are multiple settings on my iPhone that have to be correct, sometimes I have to reconnect if I turn off the car, and a few times it just refuses to tell me I have an incoming text. High-end customers demand simplicity and ease of use. This smartphone integration system is neither. The Cadillac Celestiq EV interior Image Credits:Emme HallBut at least the screens look great. Stretching across the width of the dash, there are 55 inches of high-resolution goodness. The passenger gets their own screen for streaming media and internet browsing and all climate controls are on a smaller screen below the dash. I would give my kingdom for some physical HVAC buttons, but on a car like this I’ll settle for a separate screen where at least I don’t have to scroll through a bunch of menus just to turn on the air-conditioning. Massaging seats are here, of course, and they seem to stay on a bit longer than in other luxury vehicles I’ve driven. There is no heated and cooled cup holder, an omission I think is tantamount to forgetting to serve crème fraîche with your caviar, but Cadillac told me it could add the feature if someone really wanted it. I dig the glass roof that can dim up to 20% opacity in any of the four zones. Each person in the car can customize how much light they want to let into their part of the vehicle. Image Credits:Emme HallI’m not 100% sold on the power doors, as my mind just goes to the worst-case scenario of being stuck inside the car in the Whole Foods parking lot or something. Still, it’s kind of cool to just press the brake pedal and have my door close automatically. There is also an icon on the lower screen that will close the doors, so passengers can get in on the high-tech, low-effort bandwagon. Bespoke Cadillac EV experience Image Credits:CadillacEach Celestiq will be made by hand at General Motors’ Global Technical Center in Michigan. I got a little taste of what customers will experience as I worked with a designer to select my favorite colors and materials. Choosing from what might have been 50 shades of everything was a bit overwhelming, but I walked out with my dream Celestiq in Kingfisher Tricoat, a brilliant blue that I think goes well with the long wheelbase and fastback profile, and an interior of Sheer Gray and Bahia Orange with Phantom Blue Accents. What can I say? I dig blue and orange. As the day progressed, I was treated to the Cadillac concierge service that customers will experience as they go through the buying process. Someone made sure I always had a cold Diet Dr Pepper in my hand and lunch options were tailored to my immature culinary tastes. In other words, I had tater tots for lunch and I’m not even embarrassed about it. While Celestiq customers will likely use their concierge to field logistics requests and not request copious amounts of carbs and caffeine, the point is that they will have someone to cater to their every car-buying whim. There isn’t too much out there in the uber-luxury EV market. The Cadillac Celestiq’s primary competition might be the Rolls-Royce Spectre EV with its cool starry-night headliner. Still, the Celestiq mostly sits alone in the ultra-luxury, battery-electric sector. Bentley doesn’t have a full battery-electric offering yet, just hybrids, and even Mercedes hasn’t fully electrified the Maybach S-Class sedan, although you can get a Maybach EQS SUV. If you want one in your garage, well, good luck. Cadillac will produce a limited run of 25 vehicles for 2025 and nearly all are spoken for. The company won’t give specific numbers for 2026, only saying that production will be capped at less than two Celestiqs a day.

OpenAI said on Monday it is pursuing a new restructuring plan after conversations with Delaware’s and California’s attorneys general, both of whom were closely watching as OpenAI tried to break free of its odd corporate structure. Currently, OpenAI’s nonprofit board governs the organization’s for-profit operations. Under the new plan, OpenAI’s for-profit arm will become a public benefit corporation (PBC) but will still be controlled by OpenAI’s nonprofit. The new restructuring plan may be enough to appease regulators and OpenAI’s investors, who’ve poured billions into the company in expectation of a return someday. But it could also throw a wrench into OpenAI’s future plans, particularly if the company seeks to one day go public. The IPO route Last December, OpenAI outlined a path that would’ve allowed it to spin its for-profit arm out from under the control of its nonprofit board, which is bound by certain obligations, including a clause in its charter to ensure that artificial general intelligence (AGI) benefits all humanity. That plan went out the window on Monday. Now OpenAI intends to have its nonprofit control and be a large shareholder of the aforementioned PBC. Besides allowing OpenAI to operate more like a conventional company, a simpler structure could open the door to OpenAI raising additional capital by going public via an IPO. Given OpenAI’s scale, the huge amount of cash it burns, and the public’s massive interest, an IPO seems like something OpenAI might eventually explore. Stephen Diamond, a corporate governance professor at Santa Clara University, told TechCrunch there’s a very narrow path to OpenAI becoming a public company under its newly proposed transition plan. While nonprofits can’t go public, PBCs can. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW However, there is a question as to what OpenAI’s PBC would own were OpenAI to IPO. “My sense is there’s enormous intellectual property value at the OpenAI nonprofit level,” said Diamond in an interview. “But if the PBC doesn’t own and control the core IP, but are just licensed to use it, then what’s the IPO? That’s the challenge.” Diamond noted that we don’t know the exact details of OpenAI’s plan and that it’s unclear if it’ll even be successful in the end. In an email to TechCrunch, OpenAI spokesperson Steve Sharpe said OpenAI’s nonprofit will continue to control the company’s technology and that while OpenAI has no intention of going public at this time, an IPO would be “theoretically” possible under the proposed structure. If OpenAI’s nonprofit really is in control of the organization’s critical technology, shareholders wouldn’t have much of a say in the company’s decisions, said Rose Chan Loui, the founding executive director for UCLA’s Law Program on Philanthropy and Nonprofits. Unlike buying stock in a typical company, shareholders in OpenAI would have to know that their influence over the corporation is limited. “I think an IPO is much harder in this scenario,” said Loui in an interview with TechCrunch. Bending to pressure OpenAI has been squeezed on all ends during its attempted restructuring. Just last week, a group of former OpenAI employees asked California’s and Delaware’s attorneys general to block the startup’s conversion, claiming it was at odds with OpenAI’s charitable roots. Both attorneys general told TechCrunch that they’re reviewing OpenAI’s new plan. OpenAI’s proposal also needs to appease the company’s largest private investors, including Microsoft and SoftBank, whose multibillion-dollar investments reportedly hinge on OpenAI getting some sort of restructuring over the finish line. OpenAI’s new plan gives the company a more conventional capital structure, meaning employees, investors, and the nonprofit will hold equity directly. Microsoft has not yet given its blessing to OpenAI over the new corporate structure, Bloomberg reported on Monday. The cloud provider wants to ensure the new structure adequately protects its multibillion-dollar investment in OpenAI. It’s unclear if other key stakeholders have approved the deal. No one has put more pressure on OpenAI’s restructuring than Elon Musk. The billionaire who co-founded OpenAI and now competes with it through his AI startup xAI submitted a $97 billion takeover bid to raise the price of OpenAI’s nonprofit assets and gum up the ChatGPT maker’s for-profit transition. Musk has also made OpenAI’s restructuring a focal point in his lawsuit against the startup and Microsoft. At its core, Musk’s lawsuit accuses OpenAI of abandoning its nonprofit mission to develop AGI and distribute it broadly. Last week, a federal judge denied several of OpenAI’s motions to dismiss claims in Musk’s suit. According to Diamond, this was a modest win for Musk and may have played a role in OpenAI’s changing course. However, in a briefing with members of the press on Monday, Altman reportedly denied the suit had any impact on OpenAI’s plans. Marc Toberoff, Musk’s lead counsel in his case against OpenAI, told TechCrunch the new corporate restructuring plan “changes nothing,” implying that Musk won’t be so quick to drop the case.

Uber announced its third partnership with a Chinese autonomous vehicle company this week, revealing the ride-hail and delivery giant’s appetite for global domination in the emerging robotaxi sector. Uber said Tuesday it would work with Guangzhou-based Pony AI, which late last year went public on the Nasdaq at a $5.25 billion valuation, to launch robotaxis on the Uber platform in “a key market in the Middle East” later this year. The Pony tie-up comes a day after Uber shared plans for two other strategic deals with Chinese firms: Momenta and WeRide. Uber said it would work with Momenta to launch robotaxis on its app in Europe in 2026, and with WeRide to expand into 15 cities across the Middle East and Europe over the next five years. Uber and WeRide have already launched a commercial robotaxi service in Abu Dhabi. The deals with Pony, Momenta, and WeRide add to Uber’s growing list of more than 15 autonomous vehicle partnerships across ride-hail, delivery, and freight. Most of Uber’s partners, which include Waymo and May Mobility, are based in the U.S., with a few outliers like the U.K.’s Wayve. As Uber faces perceived threats from companies like Tesla, which intends to launch its first robotaxi service in Austin next month, federal probes for allegedly misleading subscription practices, and an expected consumer spending pullback in 2025, the company is moving fast to reassure investors that there’s still plenty of room for expansion. Pairing up with Chinese companies could be a smart way to achieve that growth. Chinese AV firms are already expanding internationally. Aside from Abu Dhabi, WeRide is operating commercial services in Beijing and France, and testing in several other markets across China and Europe. Pony offers paid robotaxi rides in three Chinese cities and recently began testing in Luxembourg. Waymo, the only AV company in the U.S. that’s operating a commercial, driverless service today, has only just begun data collection in Tokyo. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW Pony’s partnership with Uber opens up the firm to the Middle Eastern market. Uber already had a strong presence in countries like the United Arab Emirates, Saudi Arabia, and Jordan — a presence that expanded in 2019 when Uber acquired its Middle Eastern rival Careem. In a press release, Pony said the goal of its Uber’s deal is to scale deployment to additional “international markets” in the future. As with Uber’s other launches, when Pony joins the app, Uber customers will have the option to have their trip fulfilled by an AV. During the initial pilot phase, the vehicles will have a safety operator onboard until the companies’ full commercial launch. Uber’s first-quarter earnings call is this week, so we’ll likely get more color on the company’s plans for growth — including its move to acquire a controlling stake in a Turkish food delivery company — and how it plans to navigate current challenges.

Thanks to a recent ruling that ordered Apple to stop charging a 27% commission on purchases through iPhone apps, Amazon’s Kindle iOS app now has a “Get Book” button that makes it easier to buy titles. “We regularly make improvements to our apps to help ensure we are providing customers the most convenient experience possible,” Amazon spokesperson Tim Gillman told The Verge, which first reported the news. “By selecting ‘Get Book’ within the Kindle for iOS app, customers can now complete their purchase through their mobile web browser.” Prior to this change, buying titles through the Kindle or Amazon app, or even viewing their prices, wasn’t allowed. Amazon isn’t the only tech giant that’s taking advantage of the recent ruling, as Spotify started allowing users to access pricing information and external payment links last Friday. While Apple is being forced to comply with the court ruling, the tech giant filed an appeal on Monday after previously stating that it strongly disagreed with the decision.

Like many leaders in tech, Arvind Krishna, the CEO of IBM, thinks federal R&D funding for AI and related technologies should be increased — not the other way around. “We have been an extremely strong advocate — and this is both the company and me personally — on that federally funded R&D should be increased,” Krishna said in an interview with TechCrunch. “That’s the position, which is unequivocal, and we strongly believe in that.” President Donald Trump and his cabinet have sought to make deep cuts to federal grantmaking and scientific research programs in pursuit of a far different set of priorities than previous administrations. In recent weeks, a key federal office for AI research, the Directorate for Technology, Innovation and Partnerships (TIP), has been hit by steep budget reductions. Meanwhile, scores of staffers at the National Institute of Standards and Technology and the National Science Foundation (NSF), which houses TIP, have found themselves on the chopping block. In its proposed budget for fiscal year 2026, the Trump administration is calling on Congress to slash over half of the NSF’s funding, eliminating what it describes as wasteful and “woke” spending. At risk are the billions of dollars awarded to AI R&D projects by the NSF and other federal agencies, like the National Institutes of Health and the Department of Energy, which are also facing cuts. Trump has also threatened to end the CHIPS Act, a major Biden administration-era law that sought to boost domestic semiconductor production in the U.S., including the production of specialized chips for AI. The office responsible for administering CHIPS Act funds was largely gutted in early March. Tech industry groups, including the Software and Information Industry Association, have sent letters to White House AI “czar” David Sacks and Commerce Secretary Howard Lutnick warning that the reductions could hobble America’s global leadership in AI. According to the U.S. Joint Economic Committee, federally funded research yields annual returns of 25% to 40%, comparing favorably to the 15% to 27% returns achieved by top-quartile VC funds. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW “We believe that [federal R&D funding] right now [is] happening at close to historic lows in terms of percentage of GDP,” Krishna told TechCrunch, “and that more federally funded R&D has an incredibly positive effect on the economy, on economic growth, on the competitiveness of the United States, and also in investing in many technologies that are essential for our future.” IBM has been directly affected by federal cutbacks. During its Q1 earnings call in April, the company revealed that 15 of its federal contracts were canceled, amounting to $100 million in future payments. Federal contracts represent between 5% and just under 10% of IBM’s consulting practice. Krishna, who was quick to note that federal funding makes up only a fraction of IBM’s overall business, said that he’s “optimistic” the government will ramp up R&D spending within a year. “I will tell you I expect that in a year … federally funded R&D for AI, for quantum, for semiconductors — I expect it to be in as good or in a better shape than it is today,” Krishna said.

If you’re a Substack publisher who prefers not to turn on your camera while livestreaming, you’re in luck. Substack announced on Tuesday that it’s launching an audio-only livestream feature for publishers who want a lower-pressure way to connect with their followers. This is also ideal for those who wish to remain anonymous. To use the feature, publishers need to click the camera icon to turn off video when going live. They can re-enable the video at any time during the stream. It’s worth noting that listeners can still comment and subscribe just like in a regular livestream. This new capability will begin rolling out next week. The new audio-only mode comes after Substack opened livestreaming to all publishers in January, which was previously limited to select creators with over 100 paying subscribers.

Spyware maker NSO Group will have to pay more than $167 million in damages to WhatsApp for a 2019 hacking campaign against more than 1,400 users. On Tuesday, after a five-year legal battle, a jury ruled that NSO Group must pay $167,254,000 in punitive damages and around $444,719 in compensatory damages. This is a huge legal win for WhatsApp, which had asked for more than $400,000 in compensatory damages, based on the time its employees had to dedicate to remediate the attacks, investigate them, and push fixes to patch the vulnerability abused by NSO Group, as well as unspecified punitive damages. WhatsApp’s spokesperson Zade Alsawah said in a statement that “our court case has made history as the first victory against illegal spyware that threatens the safety and privacy of everyone.” Alsawah said the ruling “is an important step forward for privacy and security as the first victory against the development and use of illegal spyware that threatens the safety and privacy of everyone. Today, the jury’s decision to force NSO, a notorious foreign spyware merchant, to pay damages is a critical deterrent to this malicious industry against their illegal acts aimed at American companies and the privacy and security of the people we serve.” NSO Group’s spokesperson Gil Lainer left the door open for an appeal. “We will carefully examine the verdict’s details and pursue appropriate legal remedies, including further proceedings and an appeal,” Lainer said in a statement. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW The trial, as well as the whole lawsuit, prompted a series of revelations, such as the location of the victims of the 2019 spyware campaign, as well as the names of some of NSO Group’s customers. The ruling marks the end — pending a potential appeal — of a legal battle that started in more than five years ago, when WhatsApp filed a lawsuit against the spyware maker. The Meta-owned company accused NSO Group of accessing WhatsApp servers and exploiting an audio-calling vulnerability in the chat app to target around 1,400 people, including dissidents, human rights activists, and journalists. Contact Us Do you have more information about NSO Group, or other spyware companies? From a non-work device and network, you can contact Lorenzo Franceschi-Bicchierai securely on Signal at +1 917 257 1382, or via Telegram and Keybase @lorenzofb, or email. Will Cathcart, the head of WhatsApp, explained the lawsuit’s reasoning in a Washington Post op-ed at the time, where he said that “this should serve as a wake-up call for technology companies, governments and all Internet users. Tools that enable surveillance into our private lives are being abused, and the proliferation of this technology into the hands of irresponsible companies and governments puts us all at risk.” Last December, WhatsApp won. Judge Phyllis Hamilton, who presided over the case, ruled that NSO Group was liable for breaching federal and California hacking laws in its 2019 spyware campaign against the 1,400 WhatsApp users. The judge ruled that NSO Group was also liable for breaching WhatsApp’s terms of service, which prohibit the use of the app for malicious purposes. Cathcart celebrated the December ruling saying in an X post that it was “a huge win for privacy,” and that “surveillance companies should be on notice that illegal spying will not be tolerated.” At that point, the case moved on to a jury trial to determine what damages the spyware company owed WhatsApp, which has now concluded. John Scott-Railton, a senior researcher at Citizen Lab, where he has studied the spyware industry for more than a decade, celebrated the ruling. “This is an incredible moment for those of us who have been around since the beginning of research on mercenary spyware,” Scott-Railton told TechCrunch. “NSO makes many millions of dollars helping dictators hack people. After years of every trick and delay tactic it only took the jury a day’s deliberation to see right through to the heart of the matter: NSO’s business is based on hacking American companies…so that dictators can hack dissidents.” “The company emerges from this trial severely damaged. Aside from the huge punitive damages, the bigger impact of this case has also been a huge blow to NSO’s efforts to hide their business activities,” said Scott-Railton. This story has been updated to include comments from WhatsApp and John Scott-Railton.

Rivian said in its earnings report Tuesday it will likely deliver fewer vehicles this year than previously forecasted due to President Trump’s tariffs and other regulatory changes, making it the latest automaker to be affected by the new administration’s chaotic economic policies. The company said Tuesday it expects to deliver between 40,000 and 46,000 EVs by the end of 2025. That’s despite Rivian saying one month ago that it was still holding to its estimate of delivering 46,000 to 51,000 vehicles across this year. Rivian raised its capital expenditure guidance to between $1.8 billion and $1.9 billion due to the expected impact from tariffs. The company’s previous capex guidance was between $1.6 billion to $1.7 billion, according to its 2024 shareholder letter. Rivian’s earnings announcement comes days after both Ford and General Motors pulled their guidance for the year, citing economic uncertainty related to Trump’s tariffs. Ford said it expects the tariffs to add $2.5 billion in costs across 2025, while GM told investors it expects the impact to be around $5 billion. Rivian warned investors in February that “changes to government policies and regulations, and a challenging demand environment” could threaten demand for its vehicles. Things could only get more challenging if the Trump administration, Congress, or both decide to kill the $7,500 federal tax credit for EVs. Delivering fewer than 46,000 EVs would be a step back for the electric automaker, as the company was already tracking for its third straight year with no volume growth before the guidance cut. Rivian delivered 51,579 vehicles in 2024 and 50,122 in 2023. The company’s more affordable R2 SUV, which it expects to deliver in greater numbers, won’t come until 2026. The company said Tuesday that it was able to generate $206 million of gross profit in the first quarter of 2025 on 8,640 deliveries. It was the second straight quarter the company was able to generate gross profit. This first-quarter gross profit was particularly important because it met a contractual milestone that unlocked about $1 billion in funding from Volkswagen Group as part of a joint venture with the German automaker. While gross profit may look good on the balance sheet, net income provides a more realistic view of costs. The company reported a net income loss of $541 million in the quarter, a considerable improvement from the $1.4 billion in losses in the same year-ago period. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW Automotive revenue shrank to $922 million from $1.12 billion in the first quarter of 2024, although total revenues were up slightly year-over-year thanks to a boost from sales of the company’s software and services. Total software and services revenues for the first quarter of 2025 were $318 million, nearly a fourfold increase from the $88 million in the same period last year. Rivian credited the increase to its new vehicle electrical architecture and software development services, increased remarketing sales, and an increase in repair and maintenance services. This article was first published at 4:06 p.m. ET. It has since been updated with information from Rivian’s earnings call.

The U.S. Federal Trade Commission (FTC) on Monday released new documentation detailing its new “Rule on Unfair or Deceptive Fees.” The rule, set to take effect on May 12, prohibits hidden fees for live events, hotels, and short-term rentals. It also bans practices such as “bait-and-switch pricing” and any actions that conceal or misrepresent total prices and fees. In a newly published FAQ, the FTC offers a guide for these types of businesses, providing detailed information about pricing transparency. The rule will impact businesses, including live event ticket sellers and short-term lodging providers, like hotels, motels, Airbnb, or VRBO. Third-party platforms, resellers, and travel agents are also covered by the new regulation. (Airbnb already updated its service in advance of this new regulation to show users the total cost of their stay upfront.) According to the FTC: Live event tickets include those for concerts, sporting events, music, theater, and other live performances that audiences watch as they occur, but not pre-recorded audio or visual performances. The total price must include all known charges and fees. Sites must disclose the total price upfront in ads and other offers for live-event tickets or short-term lodging. The total price must also be more prominently displayed than any other pricing information. There should be no misrepresentation about fees and charges. Sites should provide truthful information about fees, including refund policies. Sites should avoid vague terms like “convenience fees,” “service fees,” or “processing fees.” Dynamic pricing strategies are still allowed as long as the pricing information isn’t misleading. Also included in the FTC’s new FAQ are the types of fees that can be excluded, such as taxes or government fees, shipping charges, and charges for optional goods or services people may select to buy as part of the same transaction. (Note that handling charges aren’t on this list.) However, the FTC notes that businesses must disclose that it has excluded charges from the total price before asking for payment. For example, if a business excludes shipping charges from the advertised price, it’s required to clearly state the amount and purpose of those charges. The FTC first passed the rule in December 2024, a landmark regulation that marked a significant win for consumers who have been frustrated for years about hidden fees. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW

Google will unveil a new version of its Android design language at its upcoming Google I/O developer conference, according to an event schedule posted to its website as well as an accidentally published Google blog post. The I/O session title hints at significant changes in store for Android’s design system, which will evolve from “Material 3” to “Material Design 3 Expressive.” The leaked blog post indicates that Google’s updated design will focus on creating interfaces that “connect with people on an emotional level.” Design system updates like this influence how app developers build for the platform and customize their experiences for users, affecting the overall feel of what it’s like to use an Android device. While still little is known about the specifics of how Material 3 Expressive will be implemented, the blog post offered some high-level details, like how the new system is characterized by “bold use of shape and color — creating delightful user experiences.” First introduced in 2014, Material Design is Google’s open source design system that details how to create user interfaces for Android applications, including in areas like visual, motion, and interaction design. The system’s next biggest update was revealed in May 2021 as “Material You” (also known as Material 3), offering an adaptive interface that allowed users to personalize the Android interface in new ways. For instance, if you updated your Android wallpaper, the entire Android experience would also update to match its color scheme. Material 3 Expressive may not be as big an overhaul, but an iteration on the Material You language. Image Credits:Google (via 9to5Google)The leaked blog post focused more on the research and reasoning behind the update, noting how people tended to prefer a more expressive design that improves usability when studied. This type of design draws attention to key elements and important user actions, Google noted, leading to “significant gains in performance.” Of note, the company said this type of design also makes using apps easier for older adults, essentially leveling the playing field for users of all ages. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW However, Google pointed out that expressive design isn’t necessarily a “one-size fits all solution,” and different app makers will still need to respect other well-established design patterns and standards when customizing their app’s interface. Image Credits:Google (via 9to5Google)Google’s blog post was first spotted by 9to5Google and has since been saved by the Wayback Machine. Though the company quickly pulled the post down, the I/O session titled “Build next-level UX with Material 3 Expressive” shows that Google is poised to reveal the updated design system later this month at its developer conference. According to the session’s description, developers at I/O will learn to use “new emotional design patterns” to boost engagement, usability, and desire for their product. The company also plans on sharing files and alpha code at that time, so developers can begin to experiment with the changes ahead of a public release.

With long waiting lists and rising costs in overburdened healthcare systems, many people are turning to AI-powered chatbots like ChatGPT for medical self-diagnosis. About 1 in 6 American adults already use chatbots for health advice at least monthly, according to one recent survey. But placing too much trust in chatbots’ outputs can be risky, in part because people struggle to know what information to give chatbots for the best possible health recommendations, according to a recent Oxford-led study. “The study revealed a two-way communication breakdown,” Adam Mahdi, director of graduate studies at the Oxford Internet Institute and a co-author of the study, told TechCrunch. “Those using [chatbots] didn’t make better decisions than participants who relied on traditional methods like online searches or their own judgment.” For the study, the authors recruited around 1,300 people in the U.K. and gave them medical scenarios written by a group of doctors. The participants were tasked with identifying potential health conditions in the scenarios and using chatbots, as well as their own methods, to figure out possible courses of action (e.g. seeing a doctor or going to the hospital). The participants used the default AI model powering ChatGPT, GPT-4o, as well as Cohere’s Command R+ and Meta’s Llama 3, which once underpinned the company’s Meta AI assistant. According to the authors, the chatbots not only made the participants less likely to identify a relevant health condition, but it made them more likely to underestimate the severity of the conditions they did identify. Mahdi said that the participants often omitted key details when querying the chatbots or received answers that were difficult to interpret. “[T]he responses they received [from the chatbots] frequently combined good and poor recommendations,” he added. “Current evaluation methods for [chatbots] do not reflect the complexity of interacting with human users.” Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW The findings come as tech companies increasingly push AI as a way to improve health outcomes. Apple is reportedly developing an AI tool that can dispense advice related to exercise, diet, and sleep. Amazon is exploring an AI-based way to analyze medical databases for “social determinants of health.” And Microsoft is helping build AI to triage messages to care providers sent from patients. But as TechCrunch has previously reported, both professionals and patients are mixed as to whether AI is ready for higher-risk health applications. The American Medical Association recommends against physician use of chatbots like ChatGPT for assistance with clinical decisions, and major AI companies including OpenAI warn against making diagnoses based on their chatbots’ outputs. “We would recommend relying on trusted sources of information for health care decisions,” Mahdi said. “Current evaluation methods for [chatbots] do not reflect the complexity of interacting with human users. Like clinical trials for new medications, [chatbot] systems should be tested in the real world before being deployed.”

Recraft, the startup behind a mysterious image model that beat OpenAI’s DALL-E and Midjourney on a respected industry benchmark last year, has raised a $30 million Series B round led by Accel, it exclusively told TechCrunch. Other investors in the round include Khosla Ventures and Madrona. Based in San Francisco, Recraft previously raised a $12 million Series A led by Khosla in 2024. The San Francisco-based startup says it recently passed $5 million in ARR and 4 million users. The startup caught industry attention when its model, codenamed “red_panda,” topped the Artificial Analysis benchmark last year. This was actually Recraft’s V3, which earned its name because early users kept generating images of the cute mammal, Recraft’s founder and CEO Anna Veronika Dorogush told TechCrunch. Recraft says it builds its own models from scratch and competes with other image generators like Midjourney, Adobe Firefly, Stable Diffusion and Black Forest Labs, Dorogush said. But Recraft’s AI particularly excels at generating images for brands. That means allowing them to place logos exactly where needed without extra editing, or easily generating new marketing materials like brochures and posters that comply with existing branding guidelines. It’s an area where existing image models often fall short, according to Dorogush. That puts Recraft closer to competing with design tools like Canva, which also has an AI generator for branding purposes. Recraft is also notable for having a solo female founder and CEO. Dorogush founded Recraft after years of working on machine learning at Yandex – the Russian Google competitor – along with prior stints at Google and Microsoft. Before building AI models, Dorogush worked as a professional model while she earned a math and computer science degree at one of Russia’s top universities. She ended up leaving that line of work, but says it taught her that simply working hard – like showing up to endless casting calls – wasn’t enough. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW “The biggest lesson from that time was that grinding isn’t everything,” Dorogush said. “Now when building a company, I know that to succeed, we have to be excellent at what’s mission-critical. In our case, building models is very important. So we have put all the effort into excelling at this.”

Quantum computing has long been announced as “just around the corner,” but several companies are now determined to make this a commercial reality, with the promise of solving complex problems beyond classical computers’ reach. The problems in question are wide-ranging, from medicine and cybersecurity to materials science and chemistry. But first, there are very practical problems to be solved, such as developing chips that can reliably host large numbers of qubits — short for quantum bits, the fundamental unit of information in quantum computing. As usual in a high-stakes tech race with an uncertain time horizon, tech giants such as Google and Microsoft are at the forefront. Yet startups are also set to play an important role, especially tackling bottlenecks like connectivity and error correction that are critical for scaling quantum systems. But some startups and smaller tech companies are taking on the chip challenge head-on and deserve mention alongside the more attention-grabbing efforts from industry giants. While Big Tech projects typically boast the highest qubit counts, rethinking designs from first principles and using different approaches could yield equally promising results. Here are some of the companies behind the main quantum chip projects that are worth tracking. Akhetonics Akhetonics is a German photonics startup working on an all-optical, general-purpose chip — a contrarian bet in a field where most focus on narrower applications. This bold version and its first-principles approach helped the company raise a €6 million seed funding round led by Matterwave Ventures in November 2024. Alice & Bob Alice & Bob is a French startup that raised a $104 million Series B round of funding in January 2025 to continue working toward building a “fault tolerant” quantum computer. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW Unlike companies focusing only on chips, Alice & Bob is developing a full quantum computing system. To do so, it relies on cat qubits, a type of superconducting qubit designed to reduce errors and simplify error correction. Amazon Amazon officially joined the quantum chip race in early 2025 when AWS introduced Ocelot, developed in partnership with the California Institute of Technology. While this is the company’s first quantum chip, AWS had previously launched Braket, a quantum computing service, in partnership with D-Wave, IonQ, Rigetti, and others. Atom Computing Atom Computing is a U.S. company building quantum computers with arrays of optically trapped neutral atoms. At the Microsoft Ignite 2024 conference, Microsoft and Atom Computing announced plans to launch a commercial quantum computer in 2025. D-Wave D-Wave is a quantum computing company whose latest flagship system, the Advantage2 prototype, relies on a process called quantum annealing, which uses quantum physics to find the most stable, lowest-energy arrangements of elements to figure out the best mix for a given problem. Founded in 1999 as a spinoff from the University of British Columbia in Canada, D-Wave is now a public company listed on the NYSE. EeroQ Illinois-based EeroQ is a startup betting on helium for its quantum chip design. Having raised a $7.25 million seed funding round in 2022 and received regional public support, the company made a $1.1 million commitment in September 2024 toward expanding its HQ in Chicago’s Humboldt Park. Fujitsu and RIKEN In April 2025, Fujitsu and Japanese research institution RIKEN announced having developed a 256-qubit superconducting quantum computer at the RIKEN RQC-FUJITSU Collaboration Center, up from a 64-qubit iteration in 2023. Google In December 2024, Google announced Willow, its latest quantum computing chip. Google didn’t stop at saying Willow was better than its predecessor, Sycamore, or at describing it as “a major breakthrough in quantum error correction.” In a statement that made headlines, Google Quantum AI founder Hartmut Neven wrote that Willow’s performance lent “credence to the notion that quantum computation occurs in many parallel universes.” IBM IBM’s quantum efforts include Condor, its milestone superconducting chip for scaling up to 1,121 qubits, and Heron, a 156-qubit processor focused on improved performance and lower error rates. Infleqtion Founded in 2007, Infleqtion (formerly ColdQuanta) is a US-based company developing quantum computers based on neutral atom technology. Intel Intel is developing quantum computers based on silicon spin qubits. In June 2023, it unveiled Tunnel Falls, a 12-qubit research chip, and said a next-generation quantum chip based on Tunnel Falls was expected to be released in 2024, but this hasn’t happened. IonQ IonQ is a publicly listed U.S. company developing trapped-ion quantum computers, including the IonQ Forte. After going public via a SPAC in late 2021, it acquired Canadian networking specialist Entangled Networks. IQM IQM is a Finnish startup building superconducting quantum computers and is a spinout of the Aalto University and VTT Technical Research Centre of Finland. It received funding in its early days from the Business Finland government agency, followed by additional support from the EIC Accelerator program. In 2022, IQM raised €128 million in Series A2 funding led by World Fund, adding to the €39 million led by MIG Fonds it had raised in 2020 as part of what was then its Series A. This second tranche included part of a previously announced €35 million venture loan from the European Investment Bank, as well as participation from the EIC Fund and others. Microsoft In February 2025, Microsoft introduced Majorana, a quantum chip using a topological core architecture. Microsoft previously declared that it expected to build a quantum supercomputer within 10 years. Oxford Ionics As its name suggests, Oxford Ionics is a British startup that spun out of Oxford University and focuses on trapped-ion quantum computing. It raised a £30 million Series A (approximately $36 million) in early 2023, followed by an additional £2 million (approximately €2.32 million) from the UK’s National Security Strategic Investment Fund (NSSIF) that same year. More recently, it was selected for DARPA’s Quantum Benchmarking Initiative (QBI) alongside Quantinuum. Pasqal Pasqal is a French startup taking a full-stack approach to quantum computing and betting on neutral atoms. The company emerged out of the Institut d’Optique in 2019, and its co-founders include 2022 Nobel Prize laureate for physics and professor Alain Aspect. In February 2023, Pascal raised a €100 million Series B funding round led by Singapore’s Temasek, with participation from existing investors Quantonation, the Defense Innovation Fund, Daphni, and Eni Next, as well as new investors European Innovation Council (EIC) Fund, Wa’ed Ventures, and Bpifrance (through its Large Venture Fund). PsiQuantum PsiQuantum is a quantum computing startup using photonics technology and aiming to build “a 1 million-quantum-bit machine.” In February 2025, it announced Omega, a quantum photonic chipset manufactured at GlobalFoundries in New York. In 2021, PsiQuantum landed a $450 million Series D investment led by BlackRock, which is now said to be leading a potential $750 million round into the company at a $6 billion pre-money valuation. Founded by Australian academics, it is also backed by the Australian and Queensland governments. Qilimanjaro Qilimanjaro is a Spanish startup focusing on analog quantum app-specific integrated circuits (QASICs) and taking a full-stack approach, including hardware, software, and applications. Based in Barcelona, it won the Four Years From Now startup competition at Mobile World Congress in 2024, going on to receive €1.5 million in funding from Catalonia later that year. Quandela Quandela is a French startup founded in 2017 to develop photonic quantum computers. In November 2023, Quandela raised a €50 million Series B and received support from the French government through the France 2030 Plan. Quantinuum Quantinuum is a quantum computing company formed in 2021 by the merger of Cambridge Quantum and Honeywell Quantum Solutions. Its flagship product is the H-Series of trapped-ion quantum computers. In April 2024, together with Microsoft, it announced a breakthrough in error correction. QuantWare QuantWare is a Dutch startup that came up with a proprietary 3D chip architecture, VIO, which focuses on scaling bottlenecks in quantum processing units (QPUs). It started accepting preorders for Contralto-A, its first QPU for quantum error correction, in February 2025. Founded in 2020, it is a spinout of TU Delft and its affiliated research institute, QuTech. In March 2025, and following a €6 million seed round in 2023, it announced having raised a €20 million Series A (approximately $19.27 million) — including €5 million in equity out of the €7.5 million it previously secured from the European Innovation Council (the remainder is a grant). QuEra Boston-based QuEra is betting on neutral atoms as the “best approach to achieve large-scale, fault-tolerant quantum computers.” Its products include Aquila, a 256-qubit analog neutral-atom quantum computer launched in 2022 and accessible via Amazon Braket. QuEra is notably backed by Google, which led the $230 debt round that the quantum startup raised in February 2025 — with a convertible note also backed by SoftBank, Valor, and existing investors. Rigetti Computing Founded in 2013 by Chad Rigetti, Rigetti Computing is a quantum company focusing on superconducting technology. Its range of products include Ankaa-3 and the uppcoming 336-qubit Lyra system. Just like rivals D-Wave and IonQ, Rigetti Computing is a listed company; it raised just under $200 million in VC capital before going public via a SPAC in 2021. In February 2025, it signed a strategic partnership with Taiwan-based Quanta Computer, which is set to invest $35 million and purchase shares of Rigetti, while both companies invest more than $100 million each over the next five years to accelerate the development and commercialization of superconducting quantum computing. SEEQC U.S. quantum startup SEEQC — an acronym for “scalable, energy efficient quantum computing” — is a spinout of chip company Hypres, which was itself established by former employees of IBM’s superconducting electronics division. In 2023, SEEQC announced a partnership with Nvidia to build an “all-digital, ultra-low-latency chip-to-chip link between quantum computers and GPUs.” In January 2025, SEEQC raised a $30 million round of funding co-led by Booz Allen Ventures and Japanese-European VC firm NordicNinja. Other backers include Merck’s corporate venture capital arm, M Ventures. SEEQC is leading the U.K.-supported QuPharma project to explore how quantum computing can accelerate drug discovery, in partnership with BASF and Merck. SpinQ SpinQ is a Chinese startup founded in 2018 that develops quantum computers. It claims that some of them are portable and making use of nuclear magnetic resonance (NMR). Xanadu Xanadu is a Canadian startup aiming to build quantum computers through a photonic approach. In January 2025, it introduced Aurora, a 12-qubit system, including 35 photonic chips. Founded in 2016, Xanadu raised some $275 million to date, including a $100 million Series C in November 2022 at a $1 billion valuation. This story was originally published on April 28, 2025 and will be regularly updated.

View File SocialAI v2.3 - Artificial Intelligence Based Social Media Automation & Marketing Tool (SAAS) SocialAI is an AI-powered SaaS platform designed to automate and streamline social media marketing efforts. With advanced AI algorithms, SocialAI enables businesses to schedule posts, engage with followers, and analyze performance metrics across various social media platforms, all in one place. The platform’s autopilot tool creates, curates, and posts content without the need for constant supervision. It also offers intelligent insights to refine strategies, ensuring maximum reach and ROI. Ideal for businesses looking to enhance their social media presence while saving time and resources. SocialAI script has 12+ automatic payment gateways and unlimited offline payment gateways and it has built with the latest in web development technologies, including Laravel, Inertia.js, Vue.js. It has included 2 click web installer to install the script from User Interface mode. How It Works: Connect Accounts: Sign up and link your social media accounts to the SocialAI dashboard, integrating platforms like Facebook, Instagram, Twitter, LinkedIn, and more. Customize Strategy: Set your content preferences, target audience, posting frequency, and marketing goals. Content Creation & Scheduling: SocialAI’s AI generates tailored content based on your brand’s voice and audience. You can review, modify, or approve content. The autopilot feature schedules posts at optimal times for maximum engagement. Analytics & Insights: Track performance metrics, analyze engagement, and receive actionable insights to continuously improve your strategy. Refine & Automate: Adjust your campaign based on insights, or allow SocialAI’s AI to optimize and refine content delivery for sustained growth. Script Demo: Home: https://socialai.thedevstation.com/ Admin Panel Credentials Admin Login Url: https://socialai.thedevstation.com/login Email: [email protected] Password: password User Panel Credentials Candidate Login Url: https://socialai.thedevstation.com/login Email: [email protected] Password: password Submitter ceacer Submitted 05/06/2025 Category Scripts  

Anduril announced its ninth acquisition on Monday with the purchase of Dublin’s Klas, makers of ruggedized edge computing equipment for the military and first-responders. Anduril wouldn’t reveal financial details of the deal, and the purchase is subject to regulatory approval, but the company did say that Klas employs 150 people. Relatedly, on Monday Anduril also announced a new product called Menace-T. We’ll give the company points for the interesting product name, especially for a device that’s really just a bundle of compute/network connectivity, rather than, say, a fantasy-style broadsword. (Compare the name Menace to Lockheed Martin’s C2BMC, the name for its Command, Control, Battle Management & Communications products.) Klas’ flagship product, known as Voyager, is the ruggedized family of compute and networking systems that Anduril had already been using in its other Menace command center products. Voyager had also already been integrated with Anduril’s flagship Lattice software. Lattice brings sensors and AI to devices to perform tasks like object identification. But while most of us envision a portable command system being the size of a truck — which many are — Menace-T fits into two carry cases that can be set up by one person in minutes, the company says. Its goal is to bring edge computing and communications to off-grid and/or inhospitable environments. Anduril says it’s already being used in military ground vehicles and maritime vessels. One interesting use case for Menace-T is compute/communications support for the military’s Integrated Visual Augmentation System (IVAS) VR headsets. The IVAS project was initially awarded to Microsoft in 2018 after it pitched an idea of developing ruggedized HoloLens headsets for soldiers. The project was awarded an initial $21.9 billion budget. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW But after years of technical struggles, Anduril took control of the troubled contract in February — although Microsoft remains a cloud partner. Lattice had already been added to Microsoft’s IVAS headsets, bringing computer vision AI that helped the headset detect, track, and classify objects. Now Anduril thinks that the Klas technology that powers its Menace-T product can solve some of IVAS’s other historic problems, like reliable data processing. With IVAS, “there are scenarios where those soldiers need to communicate with the tactical edge to send data, to receive data, to task autonomous systems, and that’s a place where the Klas technology can help,” Tom Keane, SVP of Engineering, said at a press conference. “Klas has already been supplying technology to IVAS for several years in that context. So we expect to do more there.” Beyond Anduril’s military objectives, there are a host of other computer vision situations that are not practical until more powerful edge computing becomes available. This ranges from automotive, to industrial — or even areas like pollution monitoring. While Anduril is clearly focusing on military and related fields like law enforcement — with all the rhetoric that implies — should it solve the edge computing issue for IVAS, there are at least as many commercial possibilities for the tech. Anduril hasn’t ruled out one day pursuing those as well. “The technology and products for Klas have many use cases: military, national security, law enforcement, autonomy and more. Anduril, together with our partners, will continue to support customers from a wide array of use cases,” Keane told TechCrunch.

Not so long ago, the idea of public tech companies emerging from Latin America seemed far-fetched, and Mercado Libre once appeared as rare and mythical as a true unicorn. Today, however, the region is home to several startups that have reached billion-dollar valuations. Some of these startups, propelled into the spotlight by cross-border expansion, are now recognized beyond their home countries, with Nubank notably going public in the U.S. Yet, there is a broader cohort of Latin American scale-ups that deserve attention; many in fintech, but not exclusively. Other important sectors include e-commerce, health tech, logistics, proptech, and SaaS. Some homegrown unicorns may currently hold “paper valuations” from rounds that were raised during the 2021 peak, but the point still stands: They are worth knowing, and many could recover alongside the market, as VC investment in Latin America demonstrated resilience in 2024. As a group, these unicorns also reflect Latin America’s multiple startup hubs. While Brazil and Mexico remain leaders in numbers, unicorns have also emerged from Argentina, Colombia, Chile, and Uruguay, further strengthening these ecosystems. Let’s take a closer look at the top Latin American unicorns by valuation — although the oldest price tags often need to be taken with a grain of salt. Rappi (2015): Valued at $5.25 billion in July 2021 Coming out of Colombia, Rappi is an on-demand delivery platform that became a super app and expanded into multiple countries. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW Its rise solidified before the pandemic: In 2019, it raised a $1 billion investment from SoftBank. But its $5.25 billion valuation was attached to the round of more than $500 million it secured in July 2021. Since then, Rappi has operated in a more challenging environment, conducting multiple rounds of layoffs and facing changing gig economy legislation in Mexico, where it now plans to invest $110 million to boost its operations. However, the company still very much hopes to IPO and hired a CFO to prepare for that endeavor after reaching break-even for the first time in late 2023. QuintoAndar (2012): Valued at $5.1 billion in August 2021 QuintoAndar is a Brazilian proptech company focused on the rental and sale of residential real estate. With commercial activities in six Latin American countries and a tech hub in Europe, it made several acquisitions and grew into a group with a headcount of more than 3,500 people. In 2021, the startup was busy on the fundraising front: Less than three months after announcing a $300 million Series E at a $4 billion valuation, QuintoAndar raised an additional $120 million at a $5.1 billion valuation. With $755 million raised to date, its cap table includes Kaszek, General Atlantic, SoftBank, and Tencent. Creditas (2012): Valued at $4.8 billion in January 2022 Creditas is a Brazilian fintech player specializing in loans, including consumer credit. Its latest round of funding was a $260 million Series F in January 2022 valuing the Brazilian lender at $4.8 billion, up from $1.75 billion in December 2020. The Series F was led by Fidelity, with participation from new and existing backers, including Kaszek Ventures, QED Investors, and SoftBank. It was extended in July 2022 at the same valuation, allowing Creditas to buy the Brazilian license of Andorran bank Andbank for some $93 million. Nuvemshop (2011): Valued at $3.1 billion in August 2021 Branded as Tiendanube in Spanish-speaking markets, Nuvemshop is a Brazilian e-commerce platform designed for SMEs and entrepreneurs to sell products and services online — or in short, “Latin America’s answer to Shopify.” Its latest known valuation of $3.1 billion comes from the $500 million Series E mega-round co-led by Insight Partners and Tiger Global Management that it raised in August 2021, only a few months after its Accel-led $90 million Series D. Wildlife Studios (2011): Valued at ~$3 billion in August 2020 Wildlife Studios is a Brazilian mobile gaming company. It was co-founded by Victor Lazarte, now also a general partner at Benchmark, the VC firm that led Wildlife Studios’ $60 million Series A in 2019 at a $1.3 billion valuation. Less than a year later, the startup reached a nearly $3 billion valuation from its Series B round. In a candid conversation onstage at Slush 2023, Lazarte said that, in retrospect, raising too much capital at too high a valuation so fast was a “mistake.” In June 2023, the company announced that former Amazon executive Peter Hill would replace Lazarte as CEO. It also conducted several rounds of layoffs. Loft (2018): Valued at $2.9 billion in April 2021 Loft is a Brazilian proptech company supported by big Silicon Valley names since its genesis in 2018. Loft’s $175 million Series C was co-led by a16z and Vulcan Capital in 2020. A $425 million Series D led by New York-based D1 Capital Partners followed in March 2021, and its extension in April 2021 valued the company at $2.9 billion. The digital real estate platform wasn’t immune to the market turn. It conducted two rounds of layoffs in 2022 but denied having raised a down round in November 2022. In 2023, after fresh funding from a sovereign fund in the Middle East” at an undisclosed valuation and another round of layoffs, it claimed to have reached break-even. Ualá (2017): Valued at $2.81 billion in March 2025 Ualá is an Argentine neobank whose fintech services include also include loans, investments, payment solutions and credit scoring technology. Founded in 2017, it became a unicorn with its $350 million Series D in 2021. Since then, it raised a $300 million Series E at a $2.75 billion post-money valuation, followed by a second close in March 2025 with an additional $66 million bringing its post-money valuation to $2.81 billion. Unico (2007): Valued at $2.6 billion in April 2022 Unico is a Brazilian ID tech startup, and one of Latin America’s largest SaaS companies. Its $2.6 billion valuation is more recent than many on this list. It came from the $100 million Series D the company raised in April 2022. The round was led by Goldman Sachs, with participation from existing investors General Atlantic and SoftBank Latin America Fund. C6 Bank (2018): Valued at $2.28 billion in December 2020 C6 Bank is a Brazilian digital bank. Unlike some competitors, it hasn’t expanded beyond Brazil, where it has more than 35 million clients. C6 was valued at $2.28 billion in December 2020, six months before JPMorgan Chase acquired 40% ownership of the neobank in 2021. After increasing its stake in 2023, it now owns 46% of C6, which had its first-ever profitable year in 2024. Kavak (2016): Valued at $2.2 billion in April 2025 With backers including General Catalyst and SoftBank, Kavak is a Mexico-based e-commerce platform to buy and sell used cars online. Once valued at $8.7 billion after a Series E round that doubled its valuation in 2021, Kavak saw its valuation slashed by $6.5 billion following expansion difficulties and layoffs. After raising a $127 million equity round and securing two $200 million debt facilities in March 2025, the company aims to position itself for a potential IPO within the next three to five years. Bitso (2014): Valued at $2.2 billion in May 2021 Bitso is a Latin American cryptocurrency exchange that also facilitates cross-border payments. In May 2021, it secured a $250 million Series C round valuing the company at $2.2 billion and co-led by Tiger Global and Coatue, with participation from new and existing investors, including Kaszek and QED. CloudWalk (2013): Valued at $2.15 billion in November 2021 Known for its InfinitePay and Jim.com brands, CloudWalk is a Brazilian payment infrastructure company (not to be confused with the Chinese facial-recognition software company by the same name). CloudWalk’s $2.15 billion valuation resulted from a $150 million Series C led by Coatue in November 2021. Since then, CloudWalk achieved its first full year of profitability in 2023 and closed 2024 with $497 million in revenue. Clip (2012): Valued at $2 billion in June 2024 Briefly known as BlitzPay and founded by former PayPal employees, Clip is the Square of Latin America, with POS devices and fintech solutions for businesses. Clip became a unicorn in 2021 following a $250 million round led by SoftBank and Viking, and maintained this status since then. The $100 million round it raised in June 2024 confirmed its $2 billion valuation as the company was “on the brink of profitability,” its CEO told Bloomberg. Loggi (2013): Valued at ~$2 billion in March 2021 Loggi is a Brazil-based logistics company known for its focus on last-mile delivery. Its backers include Monashees, Qualcomm Ventures, and SoftBank. Its latest funding round was a $205 million Series F led by CapSur Capital in March 2021 at a valuation close to $2 billion. Bets are still open on who may join the list and how the rankings might be reordered, so we’ll make sure to keep it updated. This story was originally published on April 27, 2025 and will be regularly updated.

General Fusion laid off at least 25% of its employees last week, just days after hitting a key milestone for its latest fusion demonstration device. CEO Greg Twiney posted an open letter on the company’s website Monday saying that while its new LM26 device had been able to compress a plasma — something necessary for fusion conditions — General Fusion was running short of money. He wrote, “today’s funding landscape is more challenging than ever as investors and governments navigate a rapidly shifting and uncertain political and market climate.” The 23-year-old company, which many still describe as a startup, is Canada’s leading entrant in the race for commercial fusion power. It has raised $440 million, according to PitchBook, including a $22.66 million round that closed in July. Backers include Jeff Bezos, Temasek, and BDC Capital. But that money hasn’t been enough for the company to show that its unique approach to fusion is viable. General Fusion’s plight highlights the challenges that face the fusion industry. To date, only one device has been able to hit so-called scientific breakeven, a milestone that’s significant historically, but not commercially. To hit commercial breakeven, reactors need to produce dozens of times more energy than has been demonstrated so far. The road to those milestones has proven extraordinarily costly. Though General Fusion’s tally may seem impressive, it is middle of the pack. Commonwealth Fusion Systems has raised over $2 billion, Helion has pulled in over $1 billion, and upstart Pacific Fusion has been pledged $900 million in its Series A alone. Part of General Fusion’s challenge is that its taking a different approach than many of its competitors. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW Most fusion startups follow one of two paths: magnetic confinement or inertial confinement. The former uses magnetic fields to control plasma, squeezing it until it reaches the conditions for nuclei to fuse. The latter approach typically uses lasers to compress a fuel pellet. General Fusion, on the other hand, has been trying to use steam-driven pistons to compress fusion fuel. The U.S. Navy tried something similar in the 1970s to no avail, but General Fusion thinks that modern computers could solve some of the timing problems that plagued earlier attempts. It has yet to show that’s the case, but it has said that if completed, LM26 should be able to reach scientific breakeven. Now, the company will have to raise more money — and quickly — if it wants to prove its approach is a viable competitor.

Mistral AI, the French company behind AI assistant Le Chat and several foundational models, is officially regarded as one of France’s most promising tech startups and is arguably the only European company that could compete with OpenAI. But compared to its $6 billion valuation, its global market share is still relatively low. However, the recent launch of its chat assistant on mobile app stores was met with some hype, particularly in its home country. “Go and download Le Chat, which is made by Mistral, rather than ChatGPT by OpenAI — or something else,” French president Emmanuel Macron said in a TV interview ahead of the AI Action Summit in Paris. While this wave of attention may be encouraging, Mistral AI still faces challenges in competing with the likes of OpenAI — and in doing so while keeping up with its self-definition as “the world’s greenest and leading independent AI lab.” What is Mistral AI? Mistral AI has raised significant amounts of funding since its creation in 2023 with the ambition to “put frontier AI in the hands of everyone.” While this isn’t a direct jab at OpenAI, the slogan is meant to highlight the company’s advocacy for openness in AI. Its alternative to ChatGPT, chat assistant Le Chat, is now also available on iOS and Android. It reached 1 million downloads in the two weeks following its mobile release, even grabbing France’s top spot for free downloads on the iOS App Store. This comes in addition to Mistral AI’s suite of models, which includes: Mistral Large 2, the primary large language model replacing Mistral Large. Pixtral Large, unveiled in 2024 as a new addition to the Pixtral family of multimodal models. Codestral, a generative AI model for code. “Les Ministraux,” a family of models optimized for edge devices such as phones. Mistral Saba, focused on Arabic language. In March 2025, the company introduced Mistral OCR, an optical character recognition (OCR) API that can turn any PDF into a text file to make it easier for AI models to ingest. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW Who are Mistral AI’s founders? Mistral AI’s three founders share a background in AI research at major U.S. tech companies with significant operations in Paris. CEO Arthur Mensch used to work at Google’s DeepMind, while CTO Timothée Lacroix and chief scientist officer Guillaume Lample are former Meta staffers. Co-founding advisers also include Jean-Charles Samuelian-Werve (also a board member) and Charles Gorintin from health insurance startup Alan, as well as former digital minister Cédric O, which caused controversy due to his previous role. Are Mistral AI’s models open source? Not all of them. Mistral AI differentiates its premier models, whose weights are not available for commercial purposes, from its free models, for which it provides weight access under the Apache 2.0 license. Free models include research models such as Mistral NeMo, which was built in collaboration with Nvidia that the startup open-sourced in July 2024. How does Mistral AI make money? While many of Mistral AI’s offerings are free or now have free tiers, Mistral AI plans to drive some revenue from Le Chat’s paid tiers. Introduced in February 2025, Le Chat’s Pro plan is priced at $14.99 a month. On the purely B2B side, Mistral AI monetizes its premier models through APIs with usage-based pricing. Enterprises can also license these models, and the company likely also generates a significant share of its revenue from its strategic partnerships, some of which it highlighted during the Paris AI Summit. Overall, however, Mistral AI’s revenue is reportedly still in the eight-digit range, according to multiple sources. What partnerships has Mistral AI closed? In 2024, Mistral AI entered a deal with Microsoft that included a strategic partnership for distributing its AI models through Microsoft’s Azure platform and a €15 million investment. The U.K.’s Competition and Markets Authority (CMA) swiftly concluded that the deal didn’t qualify for investigation due to its small size. However, it also sparked some criticism in the EU. In January 2025, Mistral AI signed a deal with press agency Agence France-Presse (AFP) to let Chat query the AFP’s entire text archive dating back to 1983. Mistral AI also secured strategic partnerships with France’s army and job agency, shipping giant CMA, German defense tech startup Helsing, IBM, Orange, and Stellantis. How much funding has Mistral AI raised to date? As of February 2025, Mistral AI raised around €1 billion in capital to date, approximately $1.04 billion at the current exchange rate. This includes some debt financing, as well as several equity financing rounds raised in close succession. In June 2023, and before it even released its first models, Mistral AI raised a record $112 million seed round led by Lightspeed Venture Partners. Sources at the time said the seed round — Europe’s largest ever — valued the then-one-month-old startup at $260 million. Other investors in this seed round included Bpifrance, Eric Schmidt, Exor Ventures, First Minute Capital, Headline, JCDecaux Holding, La Famiglia, LocalGlobe, Motier Ventures, Rodolphe Saadé, Sofina, and Xavier Niel. Only six months later, it closed a Series A of €385 million ($415 million at the time), at a reported valuation of $2 billion. The round was led by Andreessen Horowitz (a16z), with participation from existing backer Lightspeed, as well as BNP Paribas, CMA-CGM, Conviction, Elad Gil, General Catalyst, and Salesforce. The $16.3 million convertible investment that Microsoft made in Mistral AI as part of their partnership announced in February 2024 was presented as a Series A extension, implying an unchanged valuation. In June 2024, Mistral AI then raised €600 million in a mix of equity and debt (around $640 million at the exchange rate at the time). The long-rumored round was led by General Catalyst at a $6 billion valuation, with notable investors, including Cisco, IBM, Nvidia, Samsung Venture Investment Corporation, and others. What could a Mistral AI exit look like? Mistral is “not for sale,” Mensch said in January 2025 at the World Economic Forum in Davos. “Of course, [an IPO is] the plan.” This makes sense, given how much the startup has raised so far: Even a large sale may not provide high enough multiples for its investors, not to mention sovereignty concerns depending on the acquirer. However, the only way to definitely squash persistent acquisition rumors is to scale its revenue to levels that could even remotely justify its nearly $6 billion valuation. Either way, stay tuned. This story was originally published on February 28, 2025 and will be regularly updated.

Elon Musk has long been credited with turning Tesla into a beacon of innovation, the visionary who pushed the world into an era of electrification. But his slide into politics — specifically his alignment with President Trump and activities within the federal government — has dragged Tesla into uncertainty. For many would-be Tesla buyers, the brand has lost its progressive luster. And while Musk has long been the face of Tesla, that link has taken a negative turn as the billionaire oversees mass layoffs and the elimination of federal programs via the Department of Government Efficiency. Tesla’s disappointing first-quarter earnings, which saw profits drop 71% amid weakened sales and anti-Musk sentiment, demonstrate just how much the brand has fallen out of favor. The company reported 336,681 deliveries in the first three months of 2025, down from 495,570 in Q4 2024 and 386,810 in the first quarter of last year. That’s off the back of an unimpressive 2024, which saw fewer deliveries than the year prior amid production delays and increased EV competition. For many, Musk’s political activities have turned Tesla into a symbolic pariah. Tesla trade-ins have hit record highs, while sales in Europe have dropped. Foreign leaders are calling for boycotts and the automaker’s stock price has lost about half its value since December. All the while, protesters continue to assemble outside Tesla showrooms, dealerships, and even charging stations in an effort to accelerate the automaker’s brand devaluation and with it, they hope, Musk’s wealth. “They have a certain significant public image problem that’s really something we haven’t seen before in this industry,” Sean Tucker, lead editor at Kelley Blue Book, told TechCrunch. Even one of Tesla’s most unwavering bulls, Dan Ives, has begged Musk to focus less on politics and more on his business “or else unfortunately darker times are ahead for Tesla.” And even though Musk said last month that he’d take a step back from DOGE, he’s not giving up on it either. To combat the effect this might have on Tesla, Musk has and likely will continue to lean on his sales pitch that Tesla isn’t an automaker, it’s an AI company, one that will someday commercialize autonomous vehicles and humanoid robots. The question now is whether or not the Tesla brand can recover. Techcrunch event Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Exhibit at TechCrunch Sessions: AI Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last. Berkeley, CA | June 5 BOOK NOW Below are the moments, which TechCrunch will continue to update, that are shaping perception of the Tesla brand and may have a lasting effect on its future. Drop in European sales Elon Musk Speaks at election campaign launch of AfD, Germany’s far-right party.Image Credits:Hendrik Schmidt/picture alliance / Getty ImagesTesla’s new car sales have plunged across Europe, even as sales of electric cars made by other brands rose on the continent. In the first four months of 2025, Tesla says dropped 37.2% in Europe, and in Sweden, they dropped 81% year-over-year. Germany, traditionally one of Tesla’s largest markets and home to one of its gigafactories, has turned against Musk in a big way. A recent survey of 100,000 Germans found 94% would never purchase a Tesla, and sales data backs that up. In March 2025, Tesla’s Germany sales were down 42.5%. A slight increase from the 75% sales slump in February. Succession The Wall Street Journal reported in May that Tesla’s board began searching for Musk’s potential successor sometime around early April as Musk waded further into Washington to slash government spending. When the report came out, Tesla chair Robyn Denholm declared the WSJ’s reporting to be “absolutely false.” Either way, the report does raise questions about succession planning if Musk-associated brand devaluation continues. Musk said during Tesla’s first-quarter earnings call that he would take a step back from DOGE, but, importantly, he also said he’d continue working for the agency for the remainder of Trump’s presidency. Protests are officially a risk Image Credits:Rebecca Bellan“Tesla Takedown” protests have been spreading around the world with the goal of encouraging people to stop buying Tesla’s and to sell their Tesla stock, which protesters say will hurt Musk where he holds most of his power — in his wallet (more on those below). As of the first-quarter of this year, those protests have officially become an official risk in Tesla’s regulatory filings. The new risk factor in Tesla’s latest SEC filings warns that certain criticism “has incited protests, some escalating to violence targeting our operations, products and personnel.” Tesla’s lawyers also expanded what’s at risk, saying negative perceptions resulting from the protests, along with the broader criticism of the company, “may harm our brand and our business (including sales) and make it more difficult to raise additional funds if needed.” Tariff exposure While Tesla is the automaker best positioned to parry the blow of Trump’s automotive tariffs in the U.S., the company is still exposed due to Musk’s alliance with the president. In April, Wedbush analyst Dan Ives cut Tesla’s price target to $315 from $550, once again citing Musk’s politics creating a brand crisis for the automaker. He also noted that Musk’s association with Trump, and therefore his tariff policies, are not only affecting sales in the U.S. and Europe, but also threatening Tesla’s popularity in China, “further driv[ing] Chinese consumers to buy domestic such as BYD.” A MAGA hat on wheels A sticker reading “I bought this before Elon went crazy” is pictured on a Tesla in Berlin on February 10, 2025Image Credits:obias Schwarz / AFPWhile recent studies found that Musk’s political leanings have lost him customers in what has historically been Tesla’s base — Democrats — the shift has garnered him street cred with Republicans. One study from researchers at Northeastern University, Columbia University, and the University of Iowa found that Republicans’ likelihood to purchase a Tesla moved up from 7% before Musk’s Trump endorsement to 10.2% after. That sentiment might only increase amidst Trump’s reciprocal endorsement of Musk and Tesla, including his free commercial at the White House on March 11. “Hang onto your stock“ A Tesla Cybercab prototypeImage Credits:Getty ImagesAt an impromptu, glitchy, livestreamed Tesla all-hands presentation on the evening of March 20, Musk told his employees and investors to “hang onto [their] stock” despite it feeling “like Armageddon.” “It’s very difficult for people in the stock market, especially those that look in the rearview mirror, which is most people, to imagine a future where suddenly a 10 million vehicle fleet has five to 10 times the usefulness,” Musk said, referring to the now-defunct claim that all existing Teslas have the hardware necessary to turn into fully self-driving cars with a mere software update. He advised employees to tune out the noise and focus on building the Cybercab, which would be the new priority. The Cybercab is a two-seater robotaxi with no steering wheel or pedals. Current federal regulations don’t allow for the mass production of such vehicles because they don’t have human controls, but Musk has signaled an intent to amend those rules. He noted that the Cybercab manufacturing process would be a feat of engineering, a “high-speed consumer electronics line” that “will move so fast that…it’ll be able to produce a car, ultimately, in less than five seconds.” During the call, Musk also promised to produce about 5,000 Optimus robots this year, which he believes will be Tesla’s most valuable asset. “Moment of truth“ On the same day as the sudden all-hands, Tesla bull Dan Ives wrote in a note that Musk’s involvement in DOGE has led to a “brand tornado crisis moment for Musk and Tesla.” “If you agree or disagree with DOGE it misses the point that by Musk spending 110% of his time with DOGE (and not as Tesla CEO) since President Trump got back into the White House this has essentially turned Tesla into a political symbol…. and this is a bad thing,” Ives wrote. Molotov cocktails and vandalism A Tesla Supercharger is vandalized on January 31, 2025 in San Diego, CaliforniaImage Credits:Getty ImagesTesla showrooms, charging stations, facilities, and even privately owned vehicles have been hit with attacks. In some cases, the damage is minimal — some spray paint and a keyed car. In others, violence has ensued. Dealerships have been hit with bullets and Molotov cocktails, and charging stations have been set on fire. Musk has responded by questioning why anyone would want to hurt Tesla. Trump’s response has been more pointed. He has declared that any acts of violence — which could include spray-painting — would be treated by the federal government as domestic terrorism. “I look forward to watching the sick terrorist thugs get 20-year jail sentences for what they are doing to Elon Musk and Tesla. Perhaps they could serve them in the prisons of El Salvador, which have become so recently famous for such lovely conditions!” Trump said on his social media platform, Truth Social. As of the time of this writing, the Department of Justice has brought charges against three individuals who are allegedly responsible for violent destruction of Tesla properties. Tesla Takedowns NYPD officers stand guard as protesters demonstrate during a Tesla Takedown rally on March 29, 2025Image Credits:Getty ImagesPeaceful protests have also erupted across the country as part of the so-called Tesla Takedown movement. Protesters have been showing up at Tesla dealerships and showrooms to try to discourage people from buying Tesla cars. They also aim to encourage Tesla owners to trade in their vehicles and Tesla stockholders to sell their shares. Despite these protests being peaceful, we’re beginning to see the start of violent counter-protests — like the Florida man who drove his SUV into a crowd of protestors outside a Palm Beach Tesla dealership in March. Associating with the Proud Boys Thousands of people joined in Tesla Takedown protests the last weekend in March at showrooms across the country. The crowds drew small numbers of counter-protests in some regions that Wired reports were peopled mainly by MAGA supporters. Among them were far-right extremists, including members of the Proud Boys and armed militias. The “Roman“ salute Elon Musk gestures as he speaks during the inaugural parade, in Washington, DC, on January 20, 2025Image Credits:ANGELA WEISS / AFP / Getty ImagesWhile speaking at a rally celebrating Trump’s inauguration in January, Musk did what many have perceived to be not one, but two, Nazi salutes. The damage has only been amplified amidst Musk’s vocal support for the far-right German political party AfD, as well as his history of sharing anti-semitic posts. Shortly after the salute incident, activists projected an image of Musk doing the salute, along with the word “Heil,” onto Tesla’s Berlin gigafactory. BYD breakthroughs One of Tesla’s biggest competitors, China’s BYD, has been touting significant breakthroughs in charging. The automaker said this week that its new cars can be charged in five minutes, which would give it a huge competitive edge over Tesla and indeed any other EV maker. China is one of Tesla’s largest markets, but BYD is continuing to bully the automaker on its home turf. BYD recently unveiled its “God’s Eye” advanced driver assistance system — akin to Tesla’s Autopilot and FSD — that will be installed at no additional cost on its entire EV lineup, including its ultra-cheap $9,600 Seagull hatchback. BYD continues to impress consumers with its high-tech range of EVs and hybrids and has once again surpassed Tesla on revenue. In 2024, BYD reported $107 billion in revenue compared to Tesla’s $97.7 billion. An aging lineup and Cybertruck recalls A Tesla Cybertruck on display in a Tesla showroomImage Credits:Michael M. Santiago / Getty ImagesTesla has languished with an aging lineup for years. The Cybertruck was meant to be the automaker’s next big bet that would prove it’s still capable of delivering innovative new models. And indeed, the truck has been popular among American buyers. Still, Tesla has issued eight recalls on the Cybertruck since launching the vehicle in November 2023. Most recently, Tesla had to recall 46,000 Cybertrucks because of an exterior steel trim panel on the side of the windshield that looks like it was stuck on with super glue and can just peel off.