Former DOJ antitrust chief rejects AI antitrust exemption idea

In short: Former US antitrust chief Jonathan Kanter said AI companies do not need an antitrust exemption to work on safety, and he warned it could enable anti-competitive behavior.

What happened

Jonathan Kanter, who previously led the US Department of Justice office that enforces antitrust laws, discussed AI regulation in an interview on The Verge’s Decoder podcast.

The debate centers on calls from leading AI companies, including OpenAI, Anthropic, Google DeepMind, and Elon Musk, to slow down advanced AI development and build new rules. Some leaders have suggested an antitrust exemption, which would give companies special legal permission to coordinate with each other without being accused of illegal collusion.

Kanter said companies should be able to make safer products on their own, without making deals with rivals. He compared it to airplane makers. Boeing does not need Airbus to agree to slow down in order to stop doors from falling off planes. Each company has to fix its own safety problems.

Kanter also said certain kinds of cooperation are already allowed, like sharing information about threats through a “clearinghouse” (a shared inbox for safety warnings). He argued that antitrust law does not block that kind of safety coordination.

Why it matters

If the biggest AI companies get permission to coordinate, it could reduce competition and make it harder for smaller companies to compete. Kanter’s view is that governments should set clear “rules of the road” for safety, and companies should be held responsible when their AI systems cause harm.

Source: The Verge AI