Former DHS Official Outlines Government's Legal Tools to Regulate AI

Government powers to address AI risks

The federal government already possesses legal tools to intervene if advanced artificial intelligence systems pose serious threats, according to a former Department of Homeland Security official—though those powers may have significant limitations.

Paul Rosenzweig, who previously served at DHS, outlined the existing regulatory framework available to federal agencies in a Fox News interview. His analysis comes as policymakers debate whether new legislation is needed to govern frontier AI development or whether current authorities are sufficient.

Why it matters

As AI capabilities advance rapidly, the question of regulatory readiness has moved from theoretical to urgent. Understanding what legal mechanisms already exist—and where they fall short—is essential for business leaders planning AI deployments and policymakers weighing new rules. The gap between available tools and emerging risks will shape both innovation timelines and compliance requirements across industries.

Existing authorities and their limits

Rosenzweig's assessment suggests that while federal agencies have some power to act against AI systems that present clear dangers, the scope and applicability of those authorities remain uncertain. The interview focused on "frontier AI"—the most advanced systems at the cutting edge of capability—and whether existing legal frameworks adequately cover novel risks these technologies might introduce.

The discussion reflects broader tensions in Washington over AI governance. Some lawmakers and experts argue that decades-old statutes provide adequate foundation for oversight, while others contend that AI's unique characteristics demand purpose-built regulation.

The regulatory landscape

The interview aired as part of Fox News's "Primary Politics" series and touched on questions that have divided technology policy circles. Federal agencies including the Federal Trade Commission, the Commerce Department, and sector-specific regulators have begun asserting authority over AI systems under existing consumer protection, safety, and national security laws.

However, the boundaries of those authorities—particularly for AI systems that don't fit neatly into established regulatory categories—remain contested. Rosenzweig's perspective as a former homeland security official adds weight to ongoing debates about whether national security powers provide a backstop for AI risks that other frameworks don't address.

The details were first reported by Fox News in a video segment examining Washington's capacity to respond to AI threats. As Congress considers multiple AI-related bills and agencies issue guidance on existing rules, the question of regulatory adequacy will likely intensify heading into the 2026 midterm elections.