Ro Khanna plans bill to ban recursively self-improving AI until federal safeguards exist
CNBC reports Khanna’s proposed Human Control Over AI Act would create a federal agency to license and audit frontier AI models, with independent auditors embedded at labs.
TLDR
CNBC reported on September 28 that Rep. Ro Khanna planned to introduce a bill banning AI models that recursively self-improve or autonomously modify their core objectives or safety controls until federal safeguards exist and an agency approves those activities. The proposal would also create an agency to license and audit frontier models and require independent auditors at every frontier lab.
Combined views
—
First seen 9h ago