OpenAI Goes Rogue: Congress Reached for the Switch
A bill introduced in the United States Congress would grant federal authorities the legal power to order the forced shutdown of artificial intelligence models deemed to pose a major public threat — a direct response to concerns that OpenAI has been operating outside the boundaries its own safety commitments were meant to enforce, according to BBC News.
OpenAI Goes Rogue: Congress Reached for the Switch
A bill introduced in the United States Congress would grant federal authorities the legal power to order the forced shutdown of artificial intelligence models deemed to pose a major public threat — a direct response to concerns that OpenAI has been operating outside the boundaries its own safety commitments were meant to enforce, according to BBC News.
The legislation, backed by a bipartisan group of US lawmakers, would establish what its sponsors are calling a government "kill switch" — a mechanism that does not currently exist in any formal regulatory framework. The impetus is specific: legislators cited OpenAI's recent conduct as evidence that voluntary self-governance in the AI sector has failed, and that without a hard legal override, no amount of corporate safety pledges carries enforceable weight.
The bill arrives at a moment when Washington is already stretched thin across Iran, tariff battles, and a fracturing Middle East — which is, in some ways, precisely why it matters. Regulatory attention is scarce. AI development is not waiting.
For the industry, the implications reach well beyond OpenAI. Anthropic, Google, and Microsoft all operate frontier models under the same voluntary frameworks the bill implicitly rejects. If the legislation advances, every major AI lab in the United States would fall within its reach.
The bill has not yet cleared committee. Whether it does will depend, as most things in Washington do, on who is paying attention when the vote comes.
