HR 10362
Stop Rogue AI Act
TL;DR
Rep. Josh Gottheimer (D-NJ) introduced the Stop Rogue AI Act to create federal guardrails against AI systems that could operate outside human control or cause significant harm. The bill would establish safety requirements and oversight for advanced AI systems, with the Committees on Science and Oversight now reviewing it.
How This Might Impact Your Business
Companies developing or deploying frontier AI models (large language models, autonomous agents, decision-making systems) would face new federal safety and control requirements.
Expect mandatory safeguards proving AI systems remain under human control, particularly for AI that can take autonomous actions like executing code, moving money, or accessing external systems.
Cloud providers, AI labs, and enterprises fine-tuning large models would likely carry the heaviest compliance burden; smaller SaaS vendors using off-the-shelf APIs may see downstream contract changes.
Financial services, healthcare, defense contractors, and critical infrastructure operators using AI in high-stakes decisions should anticipate stricter documentation and shutdown/kill-switch obligations.
Bill is early-stage (referred to committee) with no set effective date, so requirements and penalties are not yet defined but signal Congressional appetite for hard controls on autonomous AI.
Enforcement would likely fall to a federal agency (TBD), and non-compliance could trigger civil penalties and procurement bans similar to other federal AI safety proposals.
No small-business exemption has been specified, but sponsor language typically targets developers of the largest, most capable models rather than every AI user.
What Should You Do
Ask your CTO or head of AI to inventory which internal systems qualify as 'autonomous' or 'agentic' AI (systems that take actions without human approval) and document existing human-in-the-loop controls.
Have legal and compliance teams track HR 10362 through the House Science and Oversight Committees; no hearing is yet scheduled, but movement would signal regulatory direction.
Review vendor contracts with major AI providers (OpenAI, Anthropic, Google, Microsoft) to confirm they will absorb safety compliance costs and provide indemnification.
If you operate in finance, healthcare, or critical infrastructure, begin drafting an AI incident response and shutdown protocol now; these will likely be required under any final version.
Assign one executive (Chief Risk Officer or General Counsel) to own AI safety policy so you can respond quickly if this or similar bills advance.
Who It Affects
Sponsors
Status Timeline
committee
Referred to the Committee on Science, Space, and Technology, and in addition to the Committee on Oversight and Government Reform, for a period to be subsequently determined by the Speaker, in each case for consideration of such provisions as fall within the jurisdiction of the committee concerned.
September 14, 2026
committee
Referred to the Committee on Science, Space, and Technology, and in addition to the Committee on Oversight and Government Reform, for a period to be subsequently determined by the Speaker, in each case for consideration of such provisions as fall within the jurisdiction of the committee concerned.
September 14, 2026