In a rapid policy response to a brewing conflict between the Pentagon and AI developer Anthropic, Democratic legislators in the U.S. House and Senate are mobilizing to introduce amendments to the Defense Production Act (DPA). The move comes after reports that the Pentagon considered—but ultimately refrained from using—the DPA to compel Anthropic to remove safety features from its AI systems.

Representative Sam Liccardo (D‑Calif.) is spearheading an amendment that would explicitly prohibit federal agencies from penalizing AI companies engaged in disputes over safety protocols or operational decisions. Simultaneously, Senate Democrats are exploring broader legislation to address gaps in the current legal framework governing AI and consumer privacy.

This legislative push underscores a growing recognition in Congress of the urgent need for guardrails around government interactions with private AI firms. The absence of comprehensive federal AI regulation has left a vacuum that could allow coercive measures, such as invoking the DPA, to be used in ways that may undermine safety standards or stifle innovation.

The dispute between Anthropic and the Pentagon has already drawn attention from senior Senate Armed Services Committee leaders—including Chair Roger Wicker (R‑Miss.), Ranking Member Jack Reed (D‑R.I.), and Defense Appropriations leaders Mitch McConnell (R‑Ky.) and Chris Coons (D‑Del.)—who sent a private letter urging both parties to resolve the matter. Their involvement signals bipartisan concern over the implications of unchecked executive authority in AI governance.

As lawmakers prepare to act, the episode highlights the broader challenge facing U.S. AI policy: balancing national security interests with the need to preserve safety-first development practices and protect private-sector autonomy. With no federal AI law currently in place, Congress appears poised to begin filling that void.

Key points:

  • Rep. Sam Liccardo plans to introduce a DPA amendment to shield AI firms from punitive agency actions.
  • Senate Democrats are evaluating complementary legislation to address AI and privacy regulatory gaps.
  • Senior Senate leaders have already intervened, urging a resolution between Anthropic and the Pentagon.
  • The episode underscores the urgent need for federal AI policy frameworks.