Sponsors (3)
The bill creates a shared, standardized national reporting framework that can improve AI flaw detection, transparency, and standards, but its voluntary nature, security risks, compliance costs, and potential for slow updates may limit effectiveness and expose users.
Researchers, developers, and companies gain a national framework and shared definitions for reporting and prioritizing AI flaws, improving cross‑sector detection and remediation.
Software teams and product makers get standardized, machine‑readable reporting formats that make flaw data interoperable and faster to analyze, helping teams fix issues sooner and reducing user‑facing failures.
Public guidance and disclosure norms increase transparency about AI risks, helping consumers, regulators, and downstream users assess system safety.
Because reporting is voluntary, companies may withhold flaw information, producing incomplete coverage that limits the database's usefulness and leaves consumers and smaller users exposed.
Storing vulnerability and flaw reports in a national database risks exposing sensitive security details if dissemination and disclosure norms are insufficient, potentially aiding attackers.
Developing standards and new reporting infrastructure through NIST may impose compliance costs and technical burdens on small developers, universities, and resource‑constrained organizations.
Based on analysis of 2 sections of legislative text.
Directs NIST (with CISA) to create a voluntary AI flaw reporting program, develop taxonomies/standards, build a national reporting infrastructure, and report to Congress in three years.
Creates a NIST-led program, in consultation with CISA, to support voluntary reporting, collection, and tracking of "artificial intelligence flaws." The program must bring together industry, academia, standards bodies, civil society, nonprofits, and federal entities to develop common definitions, taxonomies, technical standards, reporting practices, and a national database or compatible infrastructure for sharing AI flaw information. Requires NIST to design or contract for the infrastructure (including machine-readable and interoperable reporting), convene stakeholders to set norms and measurement methods, and produce a report to Congress within three years describing findings, the system, and recommendations for voluntary standardized sharing. The section establishes definitions and cross-references for terms but does not amend existing U.S. Code provisions or appropriate new funds explicitly.
Official title: To direct the Director of the National Institute of Standards and Technology to develop a program for the voluntary reporting of artificial intelligence flaws and the acceleration of detection and monitoring of such flaws, and for other purposes.
Introduced June 18, 2026 by Deborah K. Ross · Last progress June 18, 2026