AI Policy Wiki
Dashboard

Andrew Garbarino

medium confidence · updated 2026-08-05

US Representative (R-NY 02) and chairman of the House Committee on Homeland Security since July 2025. Co-leads, with John Moolenaar, the joint congressional investigation into US companies' commercial adoption of PRC-developed open-weight AI models, and frames frontier AI primarily as a cyber-defense capability question.

Andrew R. Garbarino is a US Representative for New York's 2nd Congressional District and a Republican. He has chaired the House Committee on Homeland Security since July 2025. In AI policy his committee work runs along two lines: the security risks of American commercial dependence on Chinese open-weight models, pursued jointly with the House Select Committee on China; and the use of US frontier models for cyber defense of critical infrastructure, pursued through hearings, closed-door demonstrations and roundtables rather than legislation.

Joint investigation into PRC open-weight model adoption

Garbarino co-leads, with Select Committee on China chairman John Moolenaar, a joint investigation opened in April 2026 into US companies' use of AI models developed in the People's Republic of China. The investigation targets deployers rather than developers, a different instrument from the export controls and Entity List designations the Select Committee has otherwise pursued (Open-Weight Frontier Models).

The chairmen sent letters to Anysphere and Airbnb in April 2026 and to DoorDash on July 31, 2026. The DoorDash letter states that the practical advantages of PRC-developed open-weight models "do not eliminate the need for risk-based safeguards or diminish the national security concerns associated with growing dependence on models developed by entities subject to PRC jurisdiction" (Source: homeland.house.gov). It cites a post by DoorDash founder Andy Fang describing the company delegating lower-level AI work to Kimi K2.6, a model from Beijing-based Moonshot AI (Source: qz.com).

The committee describes the investigation as examining how the PRC is working to close the AI innovation gap with leading US frontier labs "through model distillation, intellectual property theft, and other illicit means," alongside the cybersecurity, data-security and critical-infrastructure risks of deploying PRC-developed AI systems in American products and services (Source: homeland.house.gov). See House Select Committee on the Strategic Competition between the United States and the Chinese Communist Party.

Frontier models as cyber-defense capability

Garbarino has described cybersecurity as his primary concern about frontier models, and has argued for defender access to them rather than for restricting them. Speaking at a Punchbowl News event in June 2026, he said "we need access to these frontier models to be able to protect our systems," and referred to a demonstration Anthropic gave the committee of Mythos, which the committee described as one of the first live demonstrations of the model given to Congress, showing members how advanced AI locates software vulnerabilities. He said of the June executive order, "I'm very happy that the White House did something with an EO… I'm happy that they included CISA in that as well" (Source: homeland.house.gov). See EO — Promoting Advanced AI Innovation and Security (Trump, signed June 2, 2026) and AI and Cybersecurity.

The committee has run much of this work outside open hearings. Garbarino has noted that industry roundtables on AI "have not been open hearings." In June 2026 the committee hosted a bipartisan closed-door demonstration of "jailbroken" AI with presentations from the National Counterterrorism Innovation, Technology, and Education Center, whose findings the committee summarized as showing extremists increasingly turning to uncensored or unrestricted models with built-in safety guardrails removed. The Subcommittee on Cybersecurity and Infrastructure Protection held a hearing that month on AI and policy solutions for securing networks and strengthening the domestic technology base, and in December 2025 held a joint hearing on threats from AI, quantum computing and hyper-scale cloud infrastructure (Source: homeland.house.gov). On July 24, 2026 members of the Homeland Security and China Select committees took part in a war-game exercise on AI-enabled cyber threats to US critical infrastructure (Source: homeland.house.gov).

Committee chairmanship

Garbarino became chairman in July 2025. Marking one year in the post, his office recorded that the committee had advanced 20 bills passed by the House and taken testimony from more than 20 leaders of Department of Homeland Security offices and agencies, with work spanning terrorism, cyberattacks, artificial intelligence, foreign adversaries and security preparations for the FIFA World Cup (Source: garbarino.house.gov). His stated committee priorities on taking the gavel were strengthening national cybersecurity, defending against terrorism and enhancing disaster preparedness (Source: cityandstateny.com).

Relationships

Sources

  • House Committee on Homeland Security, "Chairmen Garbarino, Moolenaar Continue Joint Investigation Into Security Risks Posed by PRC Open-Weight AI Models," July 31, 2026 — homeland.house.gov
  • House Committee on Homeland Security, "'We Need to Be Able to Defend At Computer Speed': Chairman Garbarino Joins Punchbowl's 'Fly Out Day' to Talk AI and Cybersecurity," June 28, 2026 — homeland.house.gov
  • Garbarino House office, "Garbarino Marks One Year as Chairman of House Committee on Homeland Security," July 21, 2026 — garbarino.house.gov
  • House Committee on Homeland Security, "House Homeland, China Select Members Participate in War Game Exercise on AI-Enabled Cyber Threats to U.S. Critical Infrastructure," July 24, 2026 — homeland.house.gov
  • Quartz, "House lawmakers are pressing DoorDash to explain its use of a Chinese AI model," July 31, 2026 — qz.com
  • City & State New York, "What does homeland security mean for Andrew Garbarino?," September 2025 — cityandstateny.com