Breaking

OpenAI, Anthropic, Google and Meta Face NYC Council Questions on AI Liability and Catastrophic Risk

October 8, 2026 at 09:35 EDT

Policy & Regulation
AI Agents
Security

Representatives from OpenAI, Anthropic, Google and Meta testified under oath before the New York City Council on October 5, as lawmakers pressed them on who would be responsible if AI agents caused serious harm and whether the companies carried insurance for catastrophic risks. The hearing turned a fast-moving technical debate into a municipal accountability fight, with council members weighing rules on independent validation, human shutdown capability, incident reporting and private lawsuits.

Oct. 5, 2026 · NYC Council

AI labs face the liability question

OpenAI, Anthropic, Google and Meta were pressed under oath on catastrophic risk, legal responsibility and whether safety promises can become enforceable rules.

bills on the city agenda

$25,000

fixed penalty per validation violation

25% / 50%

possible complainant share of recovered proceeds

Incident-reporting clocks

The city proposal would move faster than the statewide frontier-model window.

24h

City contractors and agencies

72h

State frontier-model reports

One city block equals 24 hours; the state window is three times as long.

What supporters want

External testing, whistleblower incentives, faster reporting and a right to sue when foreseeable AI misuse causes harm.

What critics flagged

Open-weight models, distributed systems and startup burdens make enforcement harder than a simple yes-or-no safety pledge.

The accountability chain

Validate

→

Shut down

→

Report

→

Sue

Lawmakers are testing whether voluntary safety processes can be translated into public obligations.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.