Ex-DOJ antitrust chief Kanter tells Decoder why AI labs don't need an antitrust exemption for safety
The Verge AI · rss · 2026-09-19
On The Verge's Decoder podcast, former DOJ antitrust chief Jonathan Kanter—who sued Google, Apple, and Ticketmaster and helped reboot US antitrust policy alongside Lina Khan—unpacks the AI safety and regulation debate.
- Context: safety-motivated departures at Anthropic and Google DeepMind, frontier CEOs calling to "pace the frontier," and requests for antitrust exemptions to coordinate on safety—drawing accusations of regulatory capture, cartel formation, and pre-IPO cost relief.
- Kanter's verdict: under both the generous reading (labs in a prisoner's dilemma need government as an enforcer) and the cynical one (cash-burning labs want permission to slow spending before IPOs), no antitrust exemption is warranted.
- Core argument: companies don't need to coordinate to ship safe products—Boeing didn't need Airbus to fix doors falling off—and firms should be held liable when their AI agents do harm, just as for human employees. Legitimate collaboration is limited to threat-sharing clearinghouses.
- He likens the current landscape to inventing cars with no traffic lights or speed limits, and discusses how regulation interacts with competition policy and China (including open-weight models).
More from Safety
- Spain reports first breach allegedly executed by an AI agent, per weekly security roundup — TechNadu · 2026-09-19
- Rant Against Both Copyright Holders and AI Firms: 'No 200-Year Hostage Over Civilization' — TinfoilTricorn · 2026-09-19
- Security researchers blast OpenAI's "nightmarish" vulnerability disclosure process — evilsocket · 2026-09-19
- Writing anti-AI-regulation op-eds with AI: cringe now, creepy soon — moultano · 2026-09-19
- Sovereign AIs may be impossible to regulate, argue researchers — leaving 'nuisance' path open — binarybits · 2026-09-19
- Social engineering, not hacking, is ASI's likeliest escape route from secure labs — ronbodkin · 2026-09-19