Turmoil at OpenAI: what’s next for the creator of ChatGPT?

On Thursday, I had dinner with Sam Altman, a few other OpenAI executives, and a small group of reporters in San Francisco. Altman answered our questions for hours. No topic was off limits, and everything, with the exception of what was said over dessert, was on the record.

It’s uncommon to have such an extended, wide-ranging interview with a major tech CEO over a meal. But there’s nothing common about the situation Altman finds himself in. ChatGPT has quickly become one of the most widely used, influential products on earth. Now, Altman is plotting an aggressive expansion into consumer hardware, brain-computer interfaces, and social media. He’s interested in buying Chrome if the US government forces Google to sell it. Oh, and he wants to raise trillions of dollars to build data centers.

OpenAI chairman Bret Taylor says the company is abandoning its effort to switch from a nonprofit to a for-profit entity after “hearing from civic leaders and engaging in constructive dialogue with the offices of the Attorney General of Delaware and the Attorney General of California.” Both attorneys general have oversight of OpenAI’s nonprofit status and could have blocked its planned restructuring, which Elon Musk, Meta, and others have publicly protested.

Now, OpenAI’s nonprofit board — the one that briefly fired CEO Sam Altman — will continue to oversee its commercial subsidiary, which is being changed from a capped, for-profit business to a public benefit corporation (PBC) like Anthropic and xAI. Previously, investors in OpenAI’s commercial entity were capped at making 100 times their money before the rest of its profits flowed back to the nonprofit.

When you can create art just by asking for it, what might you create? If you’re anything like the internet these last few days, you’d create some lovely things — stylized family portraits, a thousand different looks for the Distracted Boyfriend meme — and some truly, deeply, unfathomably horrifying ones. But most of all, it turns out, you’d turn everything, everything into a Studio Ghibli creation.

On this episode of The Vergecast, we talk a lot about the Ghibli-ification of everything this week. The Verge’s Kylie Robison joins the show to explain OpenAI’s new image generator inside ChatGPT, which is vastly more powerful and accurate than anything the company has released before. It immediately made clear both how interesting and fun these AI tools can be, and how hugely problematic they are both legally and morally. The lawsuits seem to practically write themselves, but it’s still not at all clear where any of this is headed.

The fervor around ChatGPT’s more accessible (and more advanced) image generation capabilities has forced OpenAI to “temporarily” put a rate limit on image generation requests, according to CEO Sam Altman. “It’s super fun seeing people love images in ChatGPT, but our GPUs are melting,” he posted on X today. Altman didn’t specify what the rate limit is, but said the safeguard “hopefully” won’t need to be in place for very long as OpenAI tries to increase its efficiency in handling the avalanche of requests.

The demand crunch already caused the artificial intelligence company to push back availability of the built-in image generator for users on ChatGPT’s free tier. But apparently that measure alone wasn’t enough to ease the stress on OpenAI’s infrastructure. (Altman said free users will “soon” be able to generate up to three images per day.)

AI-generated images have made significant progress since the days of abstract renderings and glitchy amalgamations. OpenAI’s newly released “Images for ChatGPT” has an uncanny ability to nail depth, shadows, and even text. It’s unleashed a frenzy of people recreating a familiar style: Hayao Miyazaki’s work at Studio Ghibli. The art style was already ubiquitous across the internet, thanks to its comforting, soft aesthetic (just look at Lofi girl) — and now, it’s a fully automated formula.

The trend kicked off pretty wholesomely. Couples transformed portraits, pet owners generated cartoonish cats, and many people are busily Ghibli-fying their families (I’ve stuck to selfies, not wanting to share with OpenAI my siblings’ likenesses). It’s an AI-generated version of the human-drawn art commissions people offer on Etsy — you and your loved ones, in the style of your favorite anime.

OpenAI is integrating new image generation capabilities directly into ChatGPT starting today — this feature is dubbed “Images in ChatGPT.” Users can now use GPT-4o to generate images within ChatGPT itself.

This initial release focuses solely on image creation and will be available across ChatGPT Plus, Pro, Team, and Free subscription tiers. The free tier’s usage limit is the same as DALL-E, spokesperson Taya Christianson told The Verge, but added that they “didn’t have a specific number to share” and ”these may change over time based on demand.“ Per the ChatGPT FAQ, free users were previously able to generate “three images per day with DALL·E 3.” As for the fate of DALL-E, Christianson said “fans” will “still have access via a custom GPT.”

In a significant executive shuffle announced Monday, OpenAI is expanding COO Brad Lightcap’s responsibilities while CEO Sam Altman shifts his attention more toward the company’s technical direction. The news was first reported by Bloomberg.

Lightcap will now “oversee day-to-day operations,” international expansion, and manage key partnerships with tech giants like Microsoft and Apple, according to Bloomberg. OpenAI has also promoted Mark Chen to chief research officer (he was recently SVP of research) and Julia Villagra to chief people officer (she was formerly VP of people).

When AI giant OpenAI submitted its “freedom-focused” policy proposal to the White House’s AI Action Plan last Thursday, it gave the Trump administration an industry wishlist: use trade laws to export American AI dominance against the looming threat of China, loosen copyright restrictions for training data (also to fight China), invest untold billions in AI infrastructure (again: China), and stop states from smothering it with hundreds of new laws.

But specifically, one law: SB 1047, California’s sweeping, controversial, and for now, defeated AI safety bill.

More than 400 members of the entertainment industry have signed a letter pushing back on OpenAI and Google’s proposal to allow AI models to train on copyrighted content, as reported earlier by Variety. The letter claims both companies “are arguing for a special government exemption” that would allow them to “freely exploit” creative industries.

The letter, which comes in response to the Trump administration’s request for feedback on its incoming AI Action Plan, is signed by stars like Ben Stiller, Mark Ruffalo, Cynthia Eviro, Cate Blanchett, Taika Waititi, Ayo Edebiri, Aubrey Plaza, Guillermo del Toro, Natasha Lyonne, Paul McCartney, and many others.

OpenAI and Google are pushing the US government to allow their AI models to train on copyrighted material. Both companies outlined their stances in proposals published this week, with OpenAI arguing that applying fair use protections to AI “is a matter of national security.”

The proposals come in response to a request from the White House, which asked governments, industry groups, private sector organizations, and others for input on President Donald Trump’s “AI Action Plan.” The initiative is supposed to “enhance America’s position as an AI powerhouse,” while preventing “burdensome requirements” from impacting innovation.

Agents are said to be the future of AI, and now OpenAI is trying to help developers build their own. The company is releasing a new Responses API that offers building blocks for developers to create agents capable of searching the web, digging through files, and performing tasks on a computer on their behalf.

“There are some agents that we will be able to build ourselves, like Deep Research and Operator,” Olivier Godement, the head of product for the OpenAI platform, tells The Verge. “But the world is so complex, there are so many industries and use cases… and so we’re super excited to provide those foundations, those building blocks for developers to build the best agents for their use case, their needs.”

OpenAI is launching GPT-4.5 today, its newest and largest AI language model. GPT-4.5 will be available as a research preview for ChatGPT Pro users to start. OpenAI is calling the release its “most knowledgeable model yet,” but initially warned that GPT-4.5 is not a frontier model and might not perform as well as o1 or o3-mini.

GPT-4.5 will have better writing capabilities, improved world knowledge, and what OpenAI calls a “refined personality over previous models.” OpenAI says interacting with GPT 4.5 will feel more “natural,” adding that the model is better at recognizing patterns and drawing connections, making it ideal for writing, programming, and “solving practical problems.”

After her sudden departure from OpenAI last fall, ex-CTO Mira Murati vanished from public view to start something new. Now, she is ready to share some details about what she’s working on.

Her new AI startup is called Thinking Machines Lab, and while the specifics of what it plans to release are still under wraps, the company says its goal is “to make AI systems more widely understood, customizable and generally capable.” The startup also promises at least some level of public transparency by pledging to regularly publish technical research and code.

OpenAI’s board of directors has responded to Elon Musk’s bid to buy the company. In a statement on X, OpenAI chair Bret Taylor said, “OpenAI is not for sale, and the board has unanimously rejected Mr. Musk’s latest attempt to disrupt his competition.”

Musk and a coalition of backers offered to buy OpenAI for $97.4 billion earlier this week. OpenAI CEO Sam Altman quickly responded to the news on X, saying “no thank you but we will buy twitter for $9.74 billion if you want.”

OpenAI is releasing a significantly expanded version of its Model Spec, a document that defines how its AI models should behave — and is making it free for anyone to use or modify.

The new 63-page specification, up from around 10 pages in its previous version, lays out guidelines for how AI models should handle everything from controversial topics to user customization. It emphasizes three main principles: customizability; transparency; and what OpenAI calls “intellectual freedom” — the ability for users to explore and debate ideas without arbitrary restrictions. The launch of the updated Model Spec comes just as CEO Sam Altman posted that the startup’s next big model, GPT-4.5 (codenamed Orion), will be released soon.

In a dramatic escalation of Silicon Valley’s most heated AI rivalry, Elon Musk just announced that he’s leading a $97.4 billion bid to buy OpenAI’s nonprofit arm, according to The Wall Street Journal. The bid was reportedly delivered to the board Monday morning.

The offer brings together a powerful coalition of backers, including Musk’s own AI company xAI and venture heavyweights like Valor Equity Partners, Hollywood mogul Ari Emanuel, and Palantir co-founder Joe Lonsdale’s venture firm 8VC.

OpenAI just made its Super Bowl debut with a 60-second spot that positions AI alongside humanity’s greatest innovations.

The commercial traces humanity’s technological evolution through a distinctive pointillism-inspired animation style, transforming abstract dots into iconic images of progress – from early tools like fire and the wheel to modern breakthroughs like DNA sequencing and space exploration. It culminates with modern AI applications, showing ChatGPT handling everyday tasks like drafting business plans and language tutoring. The ad cost roughly $14 million for the first-half placement.

OpenAI just gave itself a full rebrand, complete with a new typeface, logo, and color palette, as explained to Wallpaper in an interview about the process behind the changes. You’ll have to look closely to spot the difference between the redrawn logo and its old one, but a side-by-side comparison shows the updated “blossom” with a slightly larger space in the center and cleaner lines.

Though the original logo was designed by OpenAI co-founders Greg Brockman and Ilya Sutskever, an in-house design team led by Veit Moeller and Shannon Jager took the reins this time around, intending to create a “more organic and more human” identity, Wallpaper reports.

OpenAI is releasing a “research preview” of an AI agent called Operator that can “go to the web to perform tasks for you,” according to a blog post. “Using its own browser, it can look at a webpage and interact with it by typing, clicking, and scrolling,” OpenAI says. It’s launching first in the US for subscribers of OpenAI’s $200 per month ChatGPT Pro tier.

Operator relies a “Computer-Using Agent” model that combines GPT-4o’s vision capabilities with “advanced reasoning through reinforcement learning” to be able to interact with GUIs, OpenAI says. “Operator can ‘see’ (through screenshots) and ‘interact’ (using all the actions a mouse and keyboard allow) with a browser, enabling it to take action on the web without requiring custom API integrations,” according to OpenAI.

Microsoft and OpenAI announced Tuesday that they have adjusted their partnership so that OpenAI can access competitors’ compute.

The new agreement “includes changes to the exclusivity on new capacity, moving to a model where Microsoft has a right of first refusal (ROFR),” Microsoft says. “To further support OpenAI, Microsoft has approved OpenAI’s ability to build additional capacity, primarily for research and training of models.”

A plan to build a system of data centers for artificial intelligence has been revealed in a White House press conference, with Masayoshi Son, Sam Altman, and Larry Ellison joining Donald Trump to announce The Stargate Project. Their companies, SoftBank, OpenAI, and Oracle (respectively), along with MGX are listed as “initial equity funders” for $500 billion in investments over the next four years, “building new AI infrastructure for OpenAI in the United States.”

According to a statement from OpenAI, “Arm, Microsoft, NVIDIA, Oracle, and OpenAI” are the initial tech partners, with a buildout “currently underway” starting in Texas as other sites across the country are evaluated. It also says that “Oracle, NVIDIA, and OpenAI will closely collaborate to build and operate this computing system.”

A major copyright lawsuit against Meta has revealed a trove of internal communications about the company’s plans to develop its Llama open-source AI models, which includes discussions about avoiding “media coverage suggesting we have used a dataset we know to be pirated.”

The messages, which were part of a series of exhibits unsealed by a California court, suggest Meta used copyrighted data when training its AI systems and worked to conceal it — as it raced to beat rivals like OpenAI and Mistral. Portions of the messages were first revealed last week.

OpenAI has laid out plans to become a for-profit company. In a blog post published on Friday, OpenAI’s board said it will replace the company’s existing structure with one that puts control into the hands of its for-profit arm.

Going into 2025, OpenAI plans to become a Public Benefit Corporation⁠ (PBC), which is a for-profit company meant to operate for the good of society. This division will “run and control OpenAI’s operations and business,” while OpenAI’s nonprofit will retain a stake in the business but lose its oversight role.