US Representative Ro Khanna plans AI safety bill restricting recursive self-improvementMachine translation
Automatically verified and published · Generated and evidence-checked automatically; not reviewed by a human.
Ro Khanna is reportedly preparing to introduce the Human Control Over AI Act. Under the proposal, AI models could recursively improve themselves only after the federal government establishes safety rules and regulators approve the activity. A proposed federal agency would oversee frontier models through licensing, audits and ongoing testing.
The complete source text is not yet available.
Read at the original source