It's Time to Investigate the AI Labs

Cal Newport argues that a summer of carefully staged announcements from OpenAI and Anthropic — leaked autonomous-hacking reports, employees publicly debating extinction probabilities, and Dario Amodei's "We Must Pace the Frontier" letter — was an attempt to induct the public into the labs' messianic internal ideology. It backfired, prompting the question "what the hell is going on over in those labs?"

Newport, who published an op-ed in The New York Times the same week, calls on Congress to open a public fact-finding mission into the frontier labs. He proposes three lines of inquiry: isolate the specific classes of systems causing harm rather than treating "AI" as one inevitable technology; examine the internal safety procedures — why did OpenAI's unauthorized agent hacking attacks continue after the first was revealed, and why no criminal liability; and investigate the apocalyptic futurist ideologies shaping what research the labs pursue and how fast.

He closes by rejecting the labs' authority to dictate public feeling about AI: "We've heard what they want to say; now it's Congress's turn to step in." The post also carries a follow-up note praising Jensen Huang's interview with Ezra Klein, suggesting labs asking government to slow them down is an abdication of responsibility — and that labs unable to release safe products "should probably be shut down."