Does AI need an antitrust exemption so it doesn't kill everyone????
Summary
Former DOJ Antitrust Chief Jonathan Cantor joins Nilay Patel to analyze the AI industry’s calls for regulation and antitrust exemptions. Cantor argues that while companies cite safety, their requests for coordination risk forming a cartel to protect valuations ahead of IPOs. He asserts that existing product liability laws should hold companies responsible for their AI agents' actions, rejecting the need for new regulatory frameworks. The discussion also addresses the "national champion" strategy against China, with Cantor advocating for domestic competition over state-backed monopolies.
Topics discussed
Sponsor and show introduction
Context: AI safety concerns and antitrust reboot
Guest intro and the state of AI regulation
Debunking the AI cartel and pacing the frontier
Product liability and basic rules of the road
Legal frameworks for AI agents and tort law
Congressional inaction and the social media lesson
Meta settlement and the erosion of public trust
Public polling and corporate messaging failures
Political dynamics and bipartisan antitrust support
Reputational hits and the desire for guardrails
Competing with China and the national champion myth
Domestic competition vs foreign policy priorities
Nvidia, circular financing, and existing monopolies
Unwinding national champion policies
Midterm politics and the end of the old era
Executive branch whims and regulatory capture
Reviewing past antitrust victories and remedies
Protecting the next inflection and open source
Final policy prescriptions and closing remarks
Listen ad-free on Castria