Openness in AI is necessary, FTC’s Lina Khan said

Foundation models like OpenAI’s GPT4 are capital intensive, requiring expensive talent, expensive computing infrastructure and massive amounts of data. “These conditions have allowed, in some cases, the largest technology companies to get a layup in the AI ​​space,” Khan said.

Officials could restrict access to critical AI resources, such as compute, Khan said. Earlier this year, the FTC launched an investigation into multibillion-dollar investments by major cloud providers in generative AI startups amid growing competition concerns. That includes Microsoft’s $13 billion investment in OpenAI, as well as Google and Amazon’s multibillion-dollar investments in Anthropic.

To fully realize AI’s potential, Khan emphasizes the importance of openness in the field. What’s needed is “a real commitment to an open philosophy across the industry, which means open markets, but also open architecture, open ecosystems, and open source software.” Open source software has historically encouraged competition because it allows researchers and engineers to pool innovation. In particular, Khan calls for AI model weights to be made publicly available.

These “open-source models” lower costs and barriers to entry, enabling innovation at the application level. But he warns against an “open-to-closed playbook,” where technology may start out open-source but then change course and become proprietary over time.

Khan also notes that the FTC has worked to make the job market more inclusive. “Some of the best engineers in America were tied up in restrictive noncompete clauses,” Khan said. “We’ve heard from startups that secured funding and went to market only to find they can’t grow because the talent pool was locked down by dominant players.”

Earlier this year, the FTC banned non-compete clauses that allow “programmers, designers, and researchers across the country to move freely from company to company with their innovative ideas and unique expertise.”

Data privacy is another priority for the FTC, Khan noted. In the past, dominant players have engaged in anticompetitive practices by changing their terms of service or privacy policies, thereby backing away from their commitments to user privacy. For example, the FTC fined Meta $5 billion in 2019 for misleading users about the protection of their data.

In the AI ​​era, Khan said both consumers and businesses are uncertain about protecting their data when using legacy models. There’s “just a general uncertainty about… whether (companies’) competitively sensitive information and proprietary information is or isn’t being passed back.”

“It’s especially important to prevent these types of things on the front end, rather than playing catch-up and cleaning up on the back end when you’ve already lost so much innovation and opportunity,” Khan said.

Antitrust regulators focus only on a “narrow set of transactions that meet a narrow set of criteria”

The FTC works closely with the Antitrust Division of the Department of Justice.

“I am an enforcer of the law, not a regulator,” Jonathan Kanter, assistant attorney general for antitrust, said at the YC event. “It is not my job to regulate the market. It is my job to enforce the law only when companies break it, to ensure that competition thrives.”

In March, the DOJ filed a lawsuit against Apple for maintaining a monopoly on the smartphone market. It also filed a lawsuit against Google, accusing the company of illegally using its dominant position to manipulate the search and search advertising market.

Tech investors are frustrated by major regulators blocking mergers and acquisitions. “We can’t sell our companies,” a New York VC recently told Business Insider.

One YC founder commented on the YC forum that many startups are counting on exiting the market through acquisitions, but added that he was concerned about the “chilling effect on acquisitions” resulting from investigations by the Department of Justice and the FTC.

But Kanter emphasizes that in the vast majority of acquisitions, the Justice Department does nothing. “We focus on cases where there are competitive bottlenecks,” which is in the single digits, he said. “We only focus on deals that have issues … that’s a narrow set of deals that meet a narrow set of criteria.”

To avoid government scrutiny, Big Tech has begun exploring new paths away from traditional mergers and acquisitions. In March, Microsoft paid $650 million in a licensing deal to Inflection AI, an emerging rival to OpenAI, to use its AI models and hire most of its employees. The former Inflection CEO now heads Microsoft’s AI products and research division. Amazon followed suit, licensing the technology and hiring executives away from AI agent startup Adept in June.

Despite this, antitrust authorities continue to pay attention to this and are opening investigations into both cases.