Future of Artificial Intelligence Innovation Act of 2026

Bill Number

S. 3952

Origin Chamber

Senate

Congress

119th Congress, Session 2

Policy Area

Science, Technology, Communications

Status

Introduced

Latest Action

2026-03-19: Star Print ordered on the bill.

Last Updated

2026-03-23T15:18:23Z

AI-Generated Summary

Purpose of the Legislation

The Future of Artificial Intelligence Innovation Act of 2026 aims to advance artificial intelligence (AI) in the United States by creating voluntary standards, metrics, and testing tools for AI systems. It supports AI research, development, and workforce training, promotes innovation across companies of all sizes, and fosters international cooperation while addressing security risks and regulatory barriers. The bill emphasizes maximizing AI's benefits for public and private stakeholders without mandating compliance.

Key Provisions

The legislation is divided into three titles, focusing on standards, research, and security.

• Title I: Voluntary AI Standards, Metrics, Evaluation Tools, Testbeds, and International Cooperation
• Definitions: Defines key terms like "AI system" (machine-based systems that generate outputs like predictions with varying autonomy), "foundation model" (AI models trained on large datasets for broad adaptability), and "testbed" (facilities for rigorous AI testing).
• Center for AI Standards and Innovation (at NIST): Establishes a center within the National Institute of Standards and Technology (NIST) to develop voluntary guidelines for assessing AI security, reliability, and performance. Functions include red-teaming (adversarial testing for vulnerabilities), blue-teaming (mitigation support), detecting synthetic content (AI-generated media like deepfakes), watermarking (embedding authenticity markers in AI outputs), and creating metrics for AI evaluation. It also examines supply chain risks (e.g., semiconductors) and AI's role in fraud detection. A consortium of academia, industry, labs, and civil society will advise quarterly, with annual reports to Congress. Prohibits access for entities controlled by certain foreign governments (e.g., covered nations like China).
• Interagency Testbed Program: NIST, Department of Energy (DOE), and National Science Foundation (NSF) will create a program for public-private partnerships to test AI systems, including security assessments for cyber threats, biological risks, and critical infrastructure. Includes a voluntary testing program for foundation models and a hackathon for vulnerabilities. Metrics will evaluate program success; sunsets after 7 years.
• Materials Science Testbed: NIST and DOE will use the program to accelerate AI-driven discovery of new materials for energy storage and manufacturing, with public-private partnerships.
• Coordination and Reporting: Ensures no duplication with existing programs; requires reimbursable use of DOE resources; mandates a progress report to Congress within 1 year.
• International Coalitions: Leads alliances with like-minded countries for AI standards alignment, data sharing, and cybersecurity, excluding China until it complies with WTO commitments. Establishes trust criteria for partners (e.g., IP protections, consensus-based standards).
• Regulatory Barriers: Directs the Comptroller General to report within 1 year on federal laws/regulations hindering AI innovation (e.g., infrastructure rules), government AI adoption, and recommendations for reforms.

• Title II: AI Research, Development, and Capacity Building Activities
• Public Data for AI Systems: Amends the National AI Initiative Act of 2020 to require the Office of Science and Technology Policy (OSTP) to prioritize and curate 20 publicly available federal datasets for AI training/evaluation, focusing on public interest areas like health care and agriculture. Considers privacy, security, and national challenges; seeks public input; report to Congress within 1 year on best practices and standards.
• Federal Grand Challenges in AI: Authorizes OSTP and agencies (e.g., NSF, Commerce, Transportation) to run prize competitions (via existing authorities) for AI breakthroughs in areas like explainability (making AI decisions understandable), energy efficiency, cybersecurity, manufacturing, border security, and science applications. Includes annual priority lists, public input, and eligibility limited to U.S. entities/individuals. Agencies must report biennially on activities; sunsets after 5 years. Comptroller General will study prize competition efficacy.

• Title III: Research Security and Other Matters
• Research Security: Requires compliance with existing laws on protecting research from foreign threats.
• Hiring Expansion: Doubles NIST's authority to hire up to 30 critical technical experts (e.g., AI specialists) without standard civil service rules; extends until 2035.
• Temporary Fellows Oversight: Mandates certifications that non-employee fellows (e.g., contractors) won't perform "inherently governmental functions" (core government duties like policymaking). Agencies must submit certifications; inspectors general conduct annual audits on usage, with reports to Congress and OMB.

Significant Changes to Existing Law

• Amends the National Institute of Standards and Technology Act to add the AI Center and expand hiring powers.
• Modifies the National AI Initiative Act of 2020 by adding sections on public data priorities and grand challenges.
• Builds on existing programs (e.g., NSF's AI Research Resource pilot, DOE labs) without altering core authorities, but adds reimbursable resource sharing and FOIA exemptions for voluntary private contributions (protecting trade secrets).
• Introduces new prohibitions on foreign access and international eligibility criteria, tying cooperation to WTO compliance for China.

Potential Impacts

• Government Agencies: Enhances NIST, DOE, NSF, and OSTP capabilities for AI testing and research, potentially improving efficiency in services like fraud detection and materials discovery. Increases coordination but adds reporting burdens; reimbursements may limit resource strains.
• Citizens: Could lead to safer, more reliable AI applications in health care, agriculture, and border security, with better detection of synthetic media to combat misinformation. Public datasets may boost AI education and small business innovation.
• International Relations: Strengthens U.S. leadership in AI standards through alliances, countering adversaries like China, but may strain ties with non-eligible nations. Promotes global interoperability while protecting U.S. IP and security.

Main Stakeholders Affected

• Federal Agencies: NIST, DOE, NSF, OSTP, Commerce, and others for implementation and testing.
• Private Sector: AI developers, deployers, users, and companies of all sizes (emphasizing small/medium enterprises) through testbeds, prizes, and standards; financial and manufacturing sectors for specific applications.
• Academia and Research Institutions: Involved in consortia, data curation, and grand challenges.
• Civil Society: Consulted on priorities, ethics, and synthetic content tools.
• International Partners: Like-minded governments for coalitions; excluded nations (e.g., China) face barriers.

Notable Legal, Constitutional, or Political Implications

• Legal: Emphasizes voluntary, consensus-based standards with no new enforcement powers, preserving free market innovation while protecting confidential data under FOIA exemptions. Audits and certifications ensure accountability for temporary workers, aligning with federal acquisition rules to avoid privatization of government functions.
• Constitutional: Supports First Amendment interests by promoting open data and public input without restricting speech; upholds national security via foreign access limits, consistent with executive powers in foreign affairs.
• Political: Bipartisan (introduced by Sens. Young, Blackburn, Hickenlooper, Cantwell); signals U.S. commitment to AI competitiveness amid global rivalry, potentially influencing trade/export policies. Sunsets encourage periodic review, but exclusions (e.g., China) may spark geopolitical debates without mandating broader regulations.

This summary was generated by AI and may contain inaccuracies. Refer to the official source document for the authoritative text.

Sponsor

Sen. Young, Todd [R-IN]

Cosponsors (3)

Sen. Cantwell, Maria [D-WA], Sen. Blackburn, Marsha [R-TN], Sen. Hickenlooper, John W. [D-CO]

Recent Actions

• 2026-03-19: Star Print ordered on the bill.
• 2026-02-26: Read twice and referred to the Committee on Commerce, Science, and Transportation.
• 2026-02-26: Introduced in Senate

Bill Versions

• Future of Artificial Intelligence Innovation Act of 2026 — issued 2026-02-26 — PDF (65 pages)