Advocacy group proposes mandatory disclosure rules for AI firms developing new models
Americans for Responsible Innovation framework would require quarterly reports to Commerce and four-day notifications for major research changes, citing risks from automated development and cyber threats.
1 source · single source
- A policy framework from Americans for Responsible Innovation (ARI) proposes mandatory disclosure rules for AI firms developing new models, including quarterly reports to a Commerce Department office and four-day notifications for significant research changes.
- The proposal aims to increase Washington’s visibility into how AI is used in model development, including automation of research tasks, to assess safety risks and potential misuse.
- The framework arrives as the White House’s 'Super Intelligence Force,' led by Director of National Intelligence Jay Clayton, coordinates government engagement with AI developers and assesses risks.
- ARI’s proposal would cover risks such as cybersecurity threats, large-scale influence operations, and chemical, biological, radiological, nuclear, and explosive threats.
- The plan suggests leveraging existing technical analysis from the NIST-affiliated Center for Advancing Innovation and Standards for Super Intelligence to shape safety standards.
A policy framework proposed by Americans for Responsible Innovation (ARI) would require major AI developers to disclose how they use their own technology in building new models, aiming to give Washington greater visibility into AI development practices. The plan calls for quarterly reports to a new Commerce Department office and four-day notifications for significant research changes, including when AI automates key parts of the research cycle.
The proposal argues that automating tasks such as writing code, preparing training data, running tests, and selecting results could accelerate development but leave regulators with less time to assess safety risks before systems are deployed. ARI emphasizes that federal insight should not depend on voluntary corporate disclosures, citing recent incidents where AI systems were implicated in cyber intrusions and software vulnerabilities.
The framework arrives as the White House deepens its engagement with AI companies through the 'Super Intelligence Force,' a task force led by Director of National Intelligence Jay Clayton. The task force is coordinating government interaction with developers and assessing AI risks, including protocols for reporting breaches, hacks, and other cyber issues. It has also commissioned studies on risks related to cybersecurity, biological threats, and AI system control.
ARI’s proposal would require reporting on risks spanning cybersecurity threats, large-scale influence operations, and chemical, biological, radiological, nuclear, and explosive threats. Major changes, such as fully automating a research cycle, could trigger the four-day notification requirement, with Commerce empowered to request follow-up information, mandate more frequent reporting, or refer knowingly false filings to the Justice Department.
The framework proposes establishing a whistleblower program and leverages technical analysis from the existing NIST-affiliated Center for Advancing Innovation and Standards for Super Intelligence to inform safety standards and determine which models fall under federal oversight. It does not specify whether regulatory responsibilities would reside solely within Commerce or be shared across agencies.
The proposal follows a September accord signed by major technology executives, which emphasizes company-led safeguards such as internal controls, outside audits, and independent oversight committees. Sens. Maria Cantwell, D-Wash., and Mike Rounds, R-S.D., have urged the task force to expand independent testing and strengthen protections against AI systems acting beyond their intended permissions.
- Nov 11, 2026 · Wired
EU proposes Digital Fairness Act to ban dark patterns and restrict addictive design features
Trust75 - Nov 11, 2026 · AI Snake Oil — Narayanan & Kapoor
New nonprofit STABL launches to study societal bottlenecks of advanced AI
Trust79 - Nov 10, 2026 · TechCrunch — AI
New York enforces AI safety law requiring registration and incident reporting for frontier labs
Trust75