Policy · 1h ago
US bill mandates kill switches for dangerous AI systems
A new bill in Congress would require AI developers to implement kill switches for powerful models. The Secretary of Homeland Security could shut down AI that poses extreme risks or resists human control. The legislation aims to prevent AI from going rogue or behaving dangerously.
Meridian48 take
The bill's vague criteria for 'dangerous' AI could lead to overreach, but it signals growing urgency around AI safety regulation.
Read the full reporting
‘Powerful AI systems can go rogue, behave in extremely dangerous ways, or even resist human intervention’: A bill requiring AI systems to have a ‘kill switch’ is now in Congress →
TechRadar
ai-safetyregulation