Latest AI News
Lovable reportedly in talks to double its valuation to $13.2B
Lovable, a Swedish vibe-coding startup, is in talks to raise $300 million at a valuation of $13.2 billion — exactly double the$6.6 billionvaluation the company achieved last December,Sifted reported. Menlo Ventures, a firm that announced its latest$3 billion fundlast month, is expected to lead the round, according to the report. The less-than-three-year-old startup hit$500 millionin annualized revenue run rate in June. Lovable’s users include founders, individual designers, and salespeople building websites and e-commerce storefronts. The company also sells its vibe-coding tool to large enterprises, including Workday, Asana, and Nvidia. Vibe coding, which allows users to build software simply by describing it, is by far the most popular and lucrative use case for AI. Other high-profile vibe-coding startups include Replit, valued at$9 billionin March, and Factory, a startup that helps enterprises develop AI agents, whichraised $150 millionat a $1.5 billion valuation in April. Meanwhile, Cursor, which offers vibe coding for developers, was acquired by SpaceX for$60 billionlast month.
View
Your gaming data could be the secret to AGI, according to this Bezos-backed startup
When it comes to achieving artificial general intelligence (AGI), large language models just don’t have what it takes. Models like ChatGPT and Claude are great at text, but they’re less skilled at understanding how things actually move through space and time — an essential skill for producing intelligence that generalizes. That gap, it turns out, might be filled by gaming data. That’s the bet behind General Intuition, a Bezos-backed, New York-based startupvalued at $2.3 billion that just closed a $320 million roundwith Coatue, Eric Schmidt, and researchers at MIT and Google DeepMind joining its list of investors. On this episode of TechCrunch’sEquitypodcast, General Intuition CEO Pim de Witte joins Rebecca Bellan to dig into why world models trained on gaming data might be the next big leap in physical AI, how the company spun out of gaming platform Medal TV, and where the ethical red lines are when your models could end up being used for defense applications. Listen to the full episode to hear more about: Subscribe to Equity onYouTube,Apple Podcasts,Overcast,Spotifyand all the casts. You also can follow Equity onXandThreads, at @EquityPod.
View
These AI startups are growing revenue at faster and faster rates
As companies old and new rush to capitalize on AI, many AI startups say that their revenue is not just growing, but also rapidly accelerating, hitting their next milestones in shorter timeframes. The following list of startups have reported a pattern of such flywheel growth. One thing to note is that the underlying metrics used by these companies differ, even if they are using the term “ARR.” Some may be referring toannualized recurring revenue (ARR),or revenue under contract from a paying customer but not yet billed. Some are referring to annualized run-rate revenue, or projecting annual income by calculating 12 months of revenue that continues at the rate of the most recent month. Others are referring to “committed ARR,” or signed contracts from customers that are not onboarded yet. In the case of Gusto, it reported actualtrailing 12-month revenue. Nevertheless, each of these startups, listed in reverse chronological order to when their ARR growth was made public, reports that their revenue growth is accelerating, however they are defining it. To be sure, there are many more fast-growing AI startups than we’re naming here, but we are limiting this list to the companies hitting revenue milestones at ever-faster rates. Mercor:On Monday, Brendan Foody, co-founder and CEO of Mercor,announcedthat the company has crossed $2 billion in gross annualized revenue as of June — just four months after reaching the $1 billion milestone. The less-than-three-year-old firm, which hires domain experts to train and refine AI models,saidthat it reached a $500 million run rate in September. Anthropic:In recent months, this model maker’s revenue has been at such a historic velocity that it has mesmerized the entire AI sector. In late May, Anthropic announced that itcrossed $47 billionin revenue run rate, a milestone that came less than two months after the company reported that its revenue run rate surpassed$30 billion. The company said it reached a$9 billionrevenue run rate in late 2025, up from a reported$4 billionin July 2025. Sierra:After reaching its first$100 millionin ARR in seven quarters, Sierra — which builds customer service AI agents for enterprises — says it took just two more quarters to add another $100 million, co-founder and CEO Bret Taylorannouncedin late May. Glean:In May,Glean announcedthat it crossed $300 million in ARR. While it took the seven-year-old enterprise AI startupnine months to doubleits ARR from $100 million to $200 million, the company says it needed just six months to grow that metric from $200 million to $300 million. Gusto:The 14-year-old HR tech startup announced in May that its revenue accelerated in each of the last five quarters. The company, which was last valued at $9.3 billion in early 2022, also reported that it surpassed$1 billionin trailing 12-month revenue. Gusto’s revenue surge shows that it’s not only AI-native companies that are seeing their top-line growth supercharged by integrating the technology. Clio:This 18-year-old provider of legal practice management software saw its revenue take off sharply after embedding AI into itsoffering in 2023. The company surpassed $200 million in ARR in mid-2024, doubled that figure by late last year, and recently announced that its ARR reached$500 million.
View
Prime Intellect raises $130M Series A to help enterprises build their own AI agents
Prime Intellect, a startup that provides computing power and specialized software tools that help companies build AI agents, has raised a $130 million Series A at a $1 billion valuation. The massive round was led by Radical Ventures, with participation from Nvidia Ventures, Intel Capital, Dell Technologies Capital, Iconiq, and a long list of angel investors who are founders of notable companies, including Aravind Srinivas (Perplexity), Aaron Levie (Box), Winston Weinberg (Harvey), Jeff Wang (Cognition), and Brendan Foody (Mercor). Founded in 2024, Prime Intellect’s goal is to give organizations capabilities to train their own agentic systems without relying on frontier AI labs. While this mission would have been hard to achieve just a few years ago, the rise of reinforcement learning techniques, which iteratively reward successful task completion and penalizes errors, can allow companies to become their “own AI lab” by refining models for specific business tasks. Although it is now possible to bypass closed AI labs, the underlying infrastructure remains so complex that most companies lack the expertise to assemble these pieces into a production-ready system. That’s where Prime Intellect comes in. The startup has developed what it calls a “full stack” for AI agent development, which includes compute access, a reinforcement learning framework, and evaluation tools. Prime Intellect’s platform functions like a marketplace, providing modular access so customers can pick and choose the specific tools they need without being locked into an all-or-nothing system. “They’ve stitched this together and built it in such a way that they’re operating at the frontier in a way that’s affordable,” said David Katz, a partner at Radical Ventures. He added that while others offer bits and pieces, Prime Intellect is unique in providing the capabilities of a top-tier AI lab as a “one-stop shop” for development. The startup’s approach has attracted customers like Ramp, Zapier, and Flapping Airplanes, who pay the startup for a hosted version of its tools. This rapid adoption has propelled the company to an annualized revenue run rate of $100 million. This growth is driven by the tangible results. For example, Ramp used Prime Intellect to build an agent that helped the fintech find answers inside spreadsheets. “The result beat the frontier models on accuracy while running at faster speeds and a fraction of the cost,” Ramp’s co-founder and co-CEO Karim Atiyeh said in a statement. Another key factor driving Prime Intellect’s growth is the recent realization by companies that building on top of frontier labs carries a number of risks. Companies increasingly don’t want to provide their proprietary information to OpenAI and Anthropic due to the risk of losing control over their data. They are also wary of depending on models that can be suddenly turned off, as happened with Anthropic’s Fable last month. “How do I know that I’m not working with a company that is going to try to replace me and generalize to what I’m doing,” Katz said. “All of these things are causing people to think, ‘How do I own my own enterprise intelligence and not have these risks’.” Prime Intellect co-founder and CEO Vincent Weisser believes enterprises are looking to move away from closed source frontier models, and his company provides the infrastructure to make that transition possible. “It shouldn’t just be a few nerds in a glass tower in San Francisco that have the capability to train AI models,” he told TechCrunch. “It should be every enterprise, every nation state.”
View
OpenAI releases new voice models for more natural live conversations
OpenAI today released new conversational models, called GPT-Live-1 and GPT-Live-1 mini, claiming that they sound more natural and can handle turn-taking better. These are full-duplex models, meaning they can speak and listen at the same time, allowing users to interrupt naturally and enabling features like live translation. The company is also replacing its current Advanced Voice Mode in ChatGPT with GPT-Live-1 mini by default. Users of paid tiers will be able to access the larger GPT-Live-1 model. The previous model combined a speech-to-text model to transcribe speech, a large language model to generate responses, and a text-to-speech model to deliver the final answer. The company said in a press briefing that the new models solve issues like interrupting users while they’re talking and not having enough intelligence to answer questions. OpenAI’s new models will send the query to its latest text models like GPT-5.5 for search, reasoning, or agentic capabilities while continuing the conversation. OpenAI also showed that the model can stay silent for a long time and absorb the context of the conversation until it’s called upon. Plus, as the new voice mode has access to newer GPT models, it can also present some information in a visual format. Other startups like Monogram,which raised $40 million in seed funding from DST and Lux Capital, are also leaning into visual responses to make assistants more interactive. The company said the new voice mode in ChatGPT is designed to have longer conversations. During the briefing, ChatGPT Voice’s product lead, Atty Eleti, said he has had 30- to 40-minute-long conversations with the voice feature during walks. OpenAI thinks that voice could be the primary interface to computing for complex work. Reports have suggested that it could launcha pair of earbuds with AI capabilities this year. However, it didn’t provide any information on hardware products. “Over time, we think this will also unlock the ability to use voice as a kind of primary interface to computing, and to manage increasingly complex long-running agentic work. The kind of amazing use cases that we see people using Codex and ChatGPT to accomplish, we think voice can be the future interface to all kinds of work,” Eleti said. OpenAI has worked on bolsteringvoice-basedfeaturesover the past few years to make ChatGPT’s voice mode sound more natural. The company said that more than 150 million people talk to ChatGPT using features like Voice and Dictation. Rivals are also attempting to make assistants more expressive. BothAppleandAmazonhave updated their assistants to be more conversational with better context handling. Startups likeSesame, founded by Oculus co-founder Brendan Iribe and Ankit Kumar, also launched AI assistants with more natural conversation while completing tasks in the background. OpenAI is moving in the same direction, aiming to let users talk to its assistant hands-free for a longer time. Despite its claim that the new voice mode sounds more natural, the company emphasized that it’s not aiming to make this an AI companion. It noted that the new models have safeguards built in to give age-appropriate responses to teens and provide resources if the conversation turns to topics like self-harm. The new voice mode still needs work. During the demo, when the company showed its live translation feature in Hindi, the assistant had a heavy American accent and spoke in Hindi that was unnatural sounding and had slightly bookish tone. The company said the new mode is optimized for “most spoken languages” but didn’t specify which ones.
View
Meta wants its AI glasses to seem less creepy. Its AI strategy says otherwise.
Meta’s AI glasses have a growing reputation as acreepy technology. The company hopes to change that opinion by announcing an update that will disable the camera if the LED light that indicates the glasses are recording has been tampered with. The move is seemingly a concession toconsumersentimentthat the glasses aren’t just fun, fashionable accessories, happily promoted by Kylie Jenner, but have serious implications for consumer privacy: They can be abused as surveillance devices. Yet, even as Meta touts the new safeguard this week, the company is also pushing products and features that ask users to surrender more of their privacy to the company. Whether that’straining its AI on your images,enabling AI features using your personal contentunless you opt out, or exploring ways tocontinuously recordor usebiometric facial recognition, Meta’s vision of the future seems to always depend on collecting more of your personal data. In itsblog postabout the new camera safety feature, the company pats itself on the back, noting that “no other kind of camera has done this and we’re proud to lead the industry effort.” However, Meta also admits that the move was necessary because some people had been using tape to cover the LED light, which had already forced Meta to adapt its tech to disable recording when the LED is blocked. Determined, those same AI glasses creeps would then use “sophisticated efforts to modify or destroy the capture LED,” Meta’s announcement explains. In other words, Meta is confirming that some people who use AI glasses have hidden agendas — namely a desire to record situations or people (oftenwomen) without their consent. Despite this, the company is reportedly testing a prototype of AI glasses that would “continuously collect audio while taking photos every few seconds,” sources recently toldFinancial Times. Meta’s blog post about the glasses feature attempts to assuage people’s fears about the devices’ privacy by answering questions like “who can see the photos and videos I take on my glasses?” Meta answers by promising, “You, and only you — unless you choose to share them.” Yet, Meta’sprivacy policyhas explained that any image you share with Meta AIcan be used to train its AI. All the while, the company is facing multipleinvestigationsandlawsuitsover Meta AI glasses privacy violations.One lawsuit comes afterMeta notably canceled a contract with an outsourced tech firm after some of its Kenyan workers alleged they had to view graphic content, likesex, nudity, and people using the toilet, while training Meta’s AI using people’s Meta AI glasses’ videos. These are hardly Meta’s first scrapes with privacy violations or safety measures, either. Arguably, Meta’s reputation on privacy has been tainted for years afternumerousleaksand lost lawsuitsaboutits alleged lack of child safety measuresanddesire for growth at all costs. There arebooks by whistleblowersdocumenting its allegedabuses, not to mention previous large-scale privacy disasters, likethe Cambridge Analytica data scandalandothers. After the 2018 Cambridge Analytica scandal, Meta now insists on itsPrivacy Progress Update page, “Since 2019, we’ve invested significantly in people, products, and technology to continue to evolve our rigorous privacy program.” Still, the company plows forward with what many people would consider privacy-violating ideas. Case in point: On the same day it announced the Meta glasses’ new safeguard, it shared thatMeta AI can now use anyone’s public Instagram photosto make AI images,unless you opt out. It also built features touse Meta AI on images in your Camera Roll you’ve never sharedand implemented suchpoor privacy controlsin its Meta AI app, leading users to essentially dox themselves byrevealing their embarrassing searches. This is the same company thatApple wouldn’t partner with due to privacy concerns, thatrecords its employees’ keystrokes to train its AI, and thatplans to sell targeted ads based on data in your AI chats. So, while an LED safeguard on AI glasses might be a necessary feature, consumers clearly still have many reasons to remain distrustful of how social media will use their images and data, especially in its broader AI plans.
View
Why this CEO thinks video games make better training data than the internet
Loading the player… When it comes to achieving artificial general intelligence (AGI), large language models just don’t have what it takes. Models like ChatGPT and Claude are great at text, but they’re less skilled at understanding how things actually move through space and time — an essential skill for producing intelligence that generalizes. That gap, it turns out, might be filled by gaming data. That’s the bet behind General Intuition, a Bezos-backed, New York-based startupvalued at $2.3 billion that just closed a $320 million roundwith Coatue, Eric Schmidt, and researchers at MIT and Google DeepMind joining its list of investors. On this episode of TechCrunch’sEquitypodcast, General Intuition CEO Pim de Witte joins Rebecca Bellan to dig into why world models trained on gaming data might be the next big leap in physical AI, how the company spun out of gaming platform Medal TV, and where the ethical red lines are when your models could end up being used for defense applications. Subscribe to Equity onYouTube,Apple Podcasts,Overcast,Spotifyand all the casts. You also can follow Equity onXandThreads, at @EquityPod.
View
Google Photos adds a new AI ‘Video Remix’ tool
Google Photos is getting a new “Video Remix” feature that can edit and transform videos in seconds, Googleannouncedon Wednesday. The feature is powered byGemini Omni,Google’s recently released model that promises to “create anything from any input.” The launch is Google’s latest push to bring more generative AI tools into its consumer apps as it continues to compete with companies like Apple, OpenAI, and Adobe. By baking AI-powered video editing into Google Photos, the tech giant is making it easier for users to edit clips with a few taps instead of relying on dedicated software, giving users another reason to stay within Google’s ecosystem. The Video Remix tool can be accessed in the “Create” tab in Google Photos, allowing you to do things like apply cinematic relighting to brighten up a dark clip, swap out a plain background for something else, or add artistic styles to videos, such as watercolor, raw sketchbook, and oil painting effects. For example, you could edit a video to make it appear that you shot it in a greenhouse, relight a video with a morning glow, or paint a video in a watercolor effect. “Creating beautiful video clips shouldn’t require professional skills or hours of editing,” Google wrote in the blog post. “Now, with Video Remix in Google Photos, you can transform ordinary videos into share-worthy moments in just a few taps.” Video Remix starts rolling out today to eligible Google AI Plus, Pro, and Ultra subscribers in the U.S., Argentina, Bangladesh, Brazil, Colombia, Egypt, India, Indonesia, Japan, Mexico, Pakistan, Philippines, South Korea, and Turkey. The feature is the latest in a series of AI-powered updates introduced to Google Photos. The app recently launchednew touch-up toolsto allow users to apply subtle edits and fixes, such as removing blemishes, refining skin texture, brightening eyes, and whitening teeth. Google also announcedan AI-powered featurethat turns photos of your clothes into a digital closet where you can create new outfit ideas and virtually try on outfits.
View
Meta's Muse Image Uses Content From Public Instagram Accounts for AI-Generated Images
Meta has introduced Muse Image, its first image generation model from Meta Superintelligence Labs, alongside a new feature that allows users to generate AI images using public Instagram photos unless account owners choose to opt out. The model is rolling out across Meta AI, Instagram and WhatsApp, and lets users tag public Instagram accounts to incorporate publicly available photos into AI-generated visuals. Meta says Instagram users can disable the feature through their account settings if they do not want their public content to be used.
View
Former OpenAI exec Kevin Weil is now on the board of Stoke Space
Kevin Weil, a veteran tech executive known for stints at Twitter, Meta, Planet Labs, and OpenAI, has joined the board ofStoke Space, a well-funded Seattle startup building reusable rockets to compete with SpaceX. “It’s real simple for me,” Stoke CEO Andy Lapsa told TechCrunch of meeting Weil when he co-founded Stoke in 2020 and soon after joined Y Combinator’swinter batch. “I came out of engineering, started a company, had no idea how to fundraise. I had no idea how Silicon Valley worked. I had no network. Kevin [an early investor in the company with his wife Elizabeth, through their fundScribble Ventures] comes with all of that background and was able to help me think about fundraising and getting the company off the ground. The two kept talking while Lapsa raised $1.34 billion — including a$510 million Series Dfunding round in 2025 — to build a rapidly reusable rocket that could fly this year. Now, the time is apparently right for Weil to join the board as a director to help continue scaling the company. Stoke declined to make Weil available for an interview, and he didn’t respond to TechCrunch’s outreach. Weil’s past work has focused on digital products and platforms, which aren’t obviously on Stoke’s roadmap. He was most recently the head of OpenAI’s efforts to accelerate scientific research,leaving the companyafter that program’s work was spread more widely across the frontier lab in April. He had previously served as OpenAI’s chief product officer from June 2024 until October 2025. Weil’s last job raises one obvious question: OpenAI’s Sam Altman wasreportedlykicking the tires on Stoke last year, contemplating an investment in his own SpaceX competitor. Could Weil be the link between the frontier AI lab and a possible partner in space? Lapsa declined to comment on “gossip and rumors” about OpenAI, saying Weil’s role was to focus on Stoke itself. Stoke is building a rocket, Nova, that is intended to be completely reusable and can be flown again and again. No one has ever done that, with SpaceX coming the closest with its enormous Starship rocket. The technological challenges of reusing a rocket — particularly its ability to survive the extreme heat of reentering the Earth’s atmosphere from space — deterred even space investors with the deepest pockets. Jeff Bezos’ Blue Origin, where Lapsa once worked, has flirted with the approach, but hasn’t prioritized it. Now, though, SpaceX’s blockbuster stock market debut — with much of its value resting onElon Musk’s promisesthat Starship will be flying operational missions this year — has proven Lapsa’s foresight. Despite many billions of dollars invested in new launch vehicles, there aren’t enough rockets to go around, and the next company to get a reasonably priced rocket flying regularly promises to make a killing. “The world is realizing that launch is still not solved,” Lapsa said. “The idea of full, rapid reuse was a little bit out there at that time…that’s now been rather normalized, and people see the inevitable now.” Notably, the idea of building distributed data centers in space to leverage solar power and escape political restrictions on Earth has captured the imagination of some venture capitalists. The key obstacle there is the cost of getting all those computer chips into orbit. Space data centers “really only make sense with full rapid reuse,” Lapsa said, which could be a key differentiator for Stoke as its rocket starts flying. Military contracts will also be key to the company’s success, and Weil has experience bridging the gap between Silicon Valley and the Department of Defense; he was one offour tech movers and shakerswho joined the U.S. Army Reserve in a bid to improve recruitment and cooperation between the Army and industry. And this isn’t his first time in the space business. Weil served as the president of Planet Labs, a satellite earth observation company, for three years as it went public 2021. Whatever Weil can add to the company’s strategy as it closes in on delivering an operational launch vehicle, though, the company has to execute. “We’ve got a good chunk of the risk behind us, we’ve got more to go,” Lapsa said. “We’ll work as hard as we can, and we’ll go when it’s ready.”
View
Ford Opens Coimbatore GCC to Strengthen Business Continuity Operations
The new 800-seat facility will support finance and business functions while tapping Coimbatore’s talent pool.
View
Enterprises Have an AI Pilot Problem. The Solution May Be Neurosymbolic AI
As AI pilots struggle to reach production across enterprise environments, one supply chain technology company is betting on combining neural AI with older, structured forms of reasoning.
View
