Sen. Adam Schiff advocates for an FDA-style AI agency
In a recent interview, California Senator Adam Schiff criticized voluntary White House safety pacts and urged Congress to establish a dedicated regulatory agency to govern rapid AI development.

California Senator Adam Schiff has called for the creation of a dedicated regulatory body, comparing it to an "FDA for artificial intelligence," to oversee the rapidly advancing technology. Speaking on the Decoder podcast, the Democratic senator criticized a recent non-binding safety agreement signed by tech executives at the White House, labeling the voluntary pledge "meaningless" and "toothless." Schiff argued that the fast-paced nature of AI development requires an empowered regulatory agency with real expertise, as Congress operates too slowly to keep up with the technology's exponential growth.
As the ranking member of the Senate Judiciary subcommittee on Intellectual Property, Schiff is actively working with Senator Curtis on legislation targeting AI training data. The proposed framework would require AI developers to disclose the copyrighted materials used to train their models. Schiff noted that many frontier labs chose to "vacuum it all up" and ignore intellectual property laws during the initial development race, leaving courts and lawmakers to address the legal fallout retroactively.
The senator also highlighted severe national security concerns, pointing to the dual-use capabilities of advanced models like OpenAI's Mythos, which can be deployed for both offensive cyberattacks and defensive operations. He expressed concern over the unbridled use of AI in military targeting and criticized the administration's alleged pressure on Anthropic. According to Schiff, officials targeted Anthropic after the safety-focused startup resisted deploying its technology for domestic mass surveillance or fully autonomous weapons systems.
For AI practitioners and developers, these legislative discussions signal a shifting liability landscape. Schiff questioned whether the broad liability protections of Section 230 should apply to AI-generated outputs, suggesting that platforms may soon face direct legal consequences for the content their models synthesize. As the Supreme Court's recent Loper Bright decision threatens to complicate federal agency oversight, developers must prepare for a highly litigious environment where compliance standards remain in flux.
This is our own summary of reporting by The Verge AI



