Lina Khan on the AI-Safety Panic
Summary
Host Charlie Warzel interviews former FTC Chair Lina Kahn about the AI safety crisis triggered by an Anthropic whistleblower. They discuss why existing laws like product liability and antitrust apply to AI, countering claims of regulatory exemption. Kahn argues that while existential risks are debated, immediate harms like surveillance pricing and defective chatbots require urgent enforcement. The conversation highlights the need for robust government oversight to prevent regulatory capture and ensure accountability in the rapidly evolving AI industry.
Topics discussed
Sponsor reads and show introduction
Background on the AI safety crisis and Lina Khan's role
Why the AI safety panic is breaking through now
The black box problem and existing legal obligations
Debunking the myth of AI legal exemption
The need for new frameworks vs. self-regulation
Legal ramifications for model misbehavior and intent
Sponsor reads
Political coalitions and the dismantling of oversight
Avoiding regulatory capture and the role of money in politics
Concrete enforcement steps and industry interconnections
Existing harms: product liability and worker impact
AI's contribution to the cost of living crisis
Evaluating CEO proposals and the risk of toothless rules
Trust in AI executives and the call for antitrust exemptions
Sponsor reads
Historical parallels: railroads, cars, and web 2.0
Closing thoughts on congressional oversight and subscription
Listen ad-free on Castria