Why OpenAI's Regulatory Nightmare Is Only Getting Started

Imagine working in the OpenAI department responsible for dealing with politicians and regulators right now—things must be so intense you could almost be working for Elon Musk. Every politician worth their salt, in countries around the world, is jumping into the AI safety fray. Case in point: On Monday, the New York City Council, that bastion of AI expertise, held a hearing to discuss the risks that AI poses to humanity.

The council called whistleblowers like former OpenAI researcher Jacob Coxon to testify, along with officials from OpenAI, Google, Anthropic and Meta Platforms (Musk’s SpaceX didn’t show, and you can’t blame them). OpenAI is also facing a U.S. Senate investigation spearheaded by Sen. Josh Hawley, not to mention a Federal Trade Commission inquiry and another one by the Australian government. The state of Florida has gone to court to make life miserable for OpenAI. The Europeans, of course, have gotten into the act with their own inquiry. And these are just the regulators!

Then there are the lawsuits. OpenAI has been dealing with lawsuits filed by publishers for a while but a new wave related to AI safety is on the horizon. Already at least one lawsuit has been filed against OpenAI—by a nonprofit—over the Hugging Face hack. That case might not go anywhere (the nonprofit’s argument why it has standing to sue—that it had to divert resources to brief regulators on the episode—seems a bit thin). More suits seem certain, against OpenAI and other firms. As a think tank official told my colleague Leo Schwartz for this piece over the weekend, “all these AI companies are going to get sued up the wazoo.”

All of this is a reminder that the future of OpenAI and Anthropic doesn’t just depend on getting costs under control and maximizing revenue. If one of these rogue agents does something truly damaging, the legal liability could be enormous. This is what makes the AI story so fascinating: We really don’t know how it ends.

Nvidia’s New Record

For a company as consistent as Nvidia has been, at least in the past few years, its stock price is remarkably volatile. Shares of the AI chip giant hit a record high on Monday, closing up 2% to $238.90, 28% up so far this year. But it has been a roller coaster.

Just six months ago, Nvidia shares were around $165. They rallied to around $235 by mid-May, before dropping back down to $190 by the end of July. Yes, it’s a stock traders’ dream.

At today’s close, Nvidia has a market capitalization of $5.76 trillion, about $900 billion bigger than Apple, the No. 2 company by market value. Just two months ago, Apple’s market cap was bigger than Nvidia’s, according to Koyfin data. If you want a chance to buy into Nvidia, but think the current price is too high, just wait. You’ll get a shot before too long.

• Reflection AI, a 2-year old Nvidia-backed startup, announced its first open-weight model, adding to a collection of U.S. competitors aiming to catch up to the performance of similarly free models from China.

• OpenAI said Monday that it will test a new visual ad format that uses images to show how products and services could fit into people’s lives. Testing begins later this month in the U.S. with an initial group of advertisers.

Today on The Information’s TITV

Check out today's episode of TITV in which Akash Pasricha speaks with Cohere's Head of AI Joelle Pineau.