Sen. Adam Schiff Discusses AI Oversight, Free Speech, and Another Bid to Impeach Trump

California Senator Adam Schiff sat down with Decoder in 2026 to discuss the state of AI oversight in Congress, the limits of voluntary commitments from tech companies, and why he believes legislation cannot wait.

Background: A Pattern of Voluntary Pledges

The interview came on the heels of President Trump gathering AI CEOs at the White House to sign a non-binding pact promising to keep AI safe. Schiff described the scene as "jarring," noting that the same executives who publicly call for regulation privately sign whatever the president places in front of them. "None of these companies want to incur the wrath of this very vengeful president," he said.

Schiff pointed to the administration's treatment of Anthropic as a case in point. The company insisted on not using AI for domestic mass surveillance or fully autonomous weapons systems. The administration's response, according to Schiff, was to try to "kill Anthropic."

The National Security Contradiction

Schiff argued that the administration invokes national security to justify nearly everything—including a triumphal arch that officials now claim must store ammunition for security reasons—yet refuses to apply that same logic to AI. He raised concerns about unbridled AI implementation at the Pentagon, including the bombing of an Iranian girls' school and what role AI may have played in target selection.

When asked about AI safety, Trump reportedly told a reporter that companies "love America" and pointed to his own brain as reassurance. Schiff's assessment: "I'm not really sure that's much of an answer."

The Legislative Landscape

Schiff sits on three Senate Judiciary subcommittees central to AI regulation: Intellectual Property, Privacy and Technology, and Antitrust. He and Senator Curtis are working to require companies to disclose what copyrighted material was used to train their models.

The core problem, Schiff said, is that companies decided to "just vacuum it all up" and "worry about the legality later." He noted that AI regulation differs fundamentally from social media regulation, which was paralyzed by First Amendment challenges. AI harms—from privacy violations to autonomous weapons—do not always involve speech.

What to Watch

Schiff supports creating a new agency to regulate AI, but warned that the Supreme Court's overturning of Chevron deference in Loper Bright means any new regulation could be immediately tied up in litigation. He was blunt about Congress's track record: it did "such a shitty job" regulating social media that it cannot afford to repeat that timeline with AI.