Two House lawmakers introduced legislation to make AI pre-deployment testing and China supply-chain controls statutory requirements.

Rep. Josh Gottheimer, Co-Chair of the House AI Commission, announced two bipartisan bills on September 18, 2026, according to his office’s press release. Both are introduced legislation only. No bill numbers, committee action, or cosponsor lists are confirmed, and the details trace to Gottheimer’s release alone.

The American AI Security Act, co-sponsored by Rep. Mike Lawler, would require developers to give the NSA full access to evaluate whether a covered frontier model could enable a serious cyberattack or a chemical, biological, or radiological weapon before deployment. Review would run on a 30-day clock with one 30-day extension, plus an expedited appeals process modeled on existing national-security courts.

Per the Congressional Research Service, Executive Order 14409 already asks developers to participate in a voluntary 30-day pre-release window. The bill would convert that voluntary posture into a mandatory statutory obligation with fixed timelines and legal appeal rights.

The China FIREWALL Act, co-sponsored by Rep. Nick LaLota, would ban Chinese-developed open-weight models from government devices and bar agencies from procuring software that relies on them. The release describes it as closing a loophole left by the DeepSeek ban already enacted in the FY26 NDAA.

For CISOs serving government, the practical takeaway is tracking model provenance. A broad ban would force full lineage audits across the software stack, rather than a single-name screen. Treat both as early-warning planning signals, not active compliance requirements.

Source: Office of Rep. Josh Gottheimer