United StatesStop Rogue AI Act
An AI governance rule binding public and private bodies.
- Obligation class
- Governance, Security
As of .
What it requires
- This is a bill introduced on and referred to committee, so it binds no one unless it is enacted; as introduced its standards, rulemaking and guidance duties fall on the National Institute of Standards and Technology, the Federal Acquisition Regulatory Council and OMB, and it would reach a contractor through revised Federal Acquisition Regulation contract terms (Section 2).
- If it is enacted and you contract to supply or deploy AI agents for a federal agency under a contract entered into after enactment, maintain a continuous, machine-readable inventory of the AI agents deployed under the contract, using standardized, vendor-agnostic naming conventions (Section 2(b)(2)(A) and (4)).
- Implement AI agent identity verification mechanisms that are cryptographically verifiable at both the network and application layers (Section 2(b)(2)(B)).
- Make AI agent discovery and verification capabilities accessible to the agency without reliance on you, and enable the agency to allow, deny, or constrain AI agent-to-AI agent and AI agent-to-information system interactions (Section 2(b)(2)(C) and (D)).
- Enable the agency to provide cryptographically verifiable provenance information to authorized entities interacting with the AI agent, and generate tamper-evident, standardized logs of material AI agent actions that are portable and accessible to the agency (Section 2(b)(2)(E) and (F)).
- If you supply an information system that AI agents interact with under a contract entered into after enactment, maintain an inventory of the AI agents that interact with it and implement cryptographically verifiable identity and provenance verification for them (Section 2(b)(3)(A) to (C)).
- The standards NIST would publish provide that organizations deploying AI agents maintain a continuous, machine-readable inventory of all AI agents and do not rely solely on self-attested or single-provider assertions to establish an AI agent's identity (Section 2(a)(4)).
These lines are LexLint's reading of the law's text, as of .
What this law does
The Stop Rogue AI Act, House bill 10362 of the 119th Congress, was introduced in the House of Representatives on by Representative Gottheimer for himself and Representative Lawler. The bill was referred to the Committee on Science, Space, and Technology and, in addition, to the Committee on Oversight and Government Reform.
The bill would require the Director of the National Institute of Standards and Technology, in coordination with the Assistant Secretary of Commerce for Communications and Information, to develop, publish and maintain, not later than 1 year after enactment and annually thereafter, standards, guidelines and best practices for the secure development, deployment and operation of artificial intelligence agents by an organization.
The standards would have to maintain the capability to continuously discover, inventory, verify, and maintain organizational control over all AI agents operating within or interacting with an organization's information systems, networks, applications, services, or digital environments.
The standards would also have to enable continuous runtime monitoring and, where appropriate, inline detection and interception of AI agent interactions with tools, data sources, information systems, and other AI agents, including detection of prompt injection, data exfiltration, anomalous tool invocation, and behavioral drift.
The standards would also have to generate and retain tamper-evident, standardized logs of material AI agent actions, and ensure that the logs are portable and accessible, as appropriate and consistent with law, to deploying organizations and authorized relying parties.
The standards would have to provide that organizations deploying AI agents maintain a continuous, machine-readable inventory of all AI agents, implement independent and cryptographically verifiable AI agent identity and trust verification mechanisms, and not rely solely on self-attested or single-provider assertions of AI agent identity.
Not later than 18 months after the publication of the standards and annually thereafter, the Federal Acquisition Regulatory Council would have to propose revisions to the Federal Acquisition Regulation to require contractors and Federal agencies procuring or deploying AI agents, or information systems that interact with AI agents, to comply with the standards.
The revisions would have to ensure that contracts for the procurement or deployment of AI agents require the contractor to maintain a continuous, machine-readable inventory of the AI agents deployed under the contract, implement cryptographically verifiable identity verification, make discovery and verification capabilities accessible to the agency without reliance on the contractor, enable the agency to allow, deny, or constrain agent interactions, enable cryptographically verifiable provenance information, and generate tamper-evident, standardized logs of material AI agent actions.
The revisions would also have to ensure that contracts for the procurement or deployment of information systems that AI agents interact with carry parallel requirements for the contractor. The procurement subsection would not apply to contracts entered into before the date of enactment.
Not later than 180 days after the proposal of those revisions and annually thereafter, the Director of the Office of Management and Budget, in coordination with the Director of the Cybersecurity and Infrastructure Security Agency, would have to issue guidance to Federal agencies on implementing the procurement requirements and on the effective usage of AI agents.
The bill defines an AI agent as a software-based system that uses an AI model to perceive, plan, or make decisions, autonomously interacts with other software systems, digital services, users, external environments, or AI agents on behalf of a person or organization, and involves minimal or no human interaction beyond its initial direction.
When LexLint raises it
When your app profile says your app distributes a software product.