Efforts to regulate artificial intelligence are gaining momentum in Washington: a bipartisan group of U.S. senators rolled out an AI transparency bill on Thursday. The measure was introduced by Sens. Chris Coons (D-Del.), Katie Britt (R-Ala.), Brian Schatz (D-Hawaii) and James Lankford (R-Okla.).
Details first shared with Semafor indicate the proposal would empower the Federal Trade Commission (FTC) to enforce disclosure requirements for certain AI companies. Under the bill, affected firms would need to provide information about how their AI models function and what safeguards they have in place to prevent misuse.
The stated aim is to give the public and users clearer insight into the safety guardrails applied to large AI models. As Sens. Coons put it: “Americans deserve to know what [AI companies] are actually doing to place safety guardrails on large AI models.”
This bill is relatively narrowly focused on transparency and awareness of AI systems’ operation and protections rather than sweeping regulatory changes. Other members of Congress have pursued more expansive proposals, ranging from outright bans on “superintelligence” to economic plans for an AI-driven economy.
Key points:
- The bipartisan sponsorship could increase the bill's prospects in a divided Congress.
- Assigning enforcement to the FTC would create a single administrative authority to oversee disclosure compliance.
- By concentrating on disclosure and safeguards, the measure is a more modest intervention compared with broader legislative proposals.
The proposal’s primary effect would be to improve public access to information about how AI systems work and how companies are mitigating associated risks.
