Scorecard

Know where Washington stands.

We evaluate important federal AI legislation against a public five-question test. Member scores will be published as recorded votes accumulate. Our goal is simple: make AI policy understandable and government accountable to the public.

We score on

National security
AI safety
American innovation
Biosecurity
Cybersecurity
Child safety
Human accountability
Workforce preparedness

01 Does it protect Americans?
02 Does it preserve American innovation?
03 Does it strengthen national security?
04 Does it maintain meaningful human accountability?
05 Will it still work as AI becomes dramatically more capable?

H.R. 9363 · House · June 25, 2026

AI Security and Innovation Act

Rep. Jay Obernolte (R-CA) and Rep. Valerie Foushee (D-NC)

Support

Authorizes a Center for AI Security and Innovation at NIST to keep the United States leading in research, development, and evaluation of AI systems.

Status: Ordered reported, 29–0

Protect: Yes
Innovate: Yes
Security: Yes
Accountable: Yes
Durable: Yes

A rare bipartisan vehicle that treats evaluation capacity as a national asset rather than a brake on research. Independent testing infrastructure is how America leads and stays safe at the same time.

AI safetyAmerican innovationNational security

Source

H.R. 8893 · House · June 25, 2026

Protecting Consumers From Deceptive AI Act

Rep. Valerie Foushee (D-NC), Rep. Don Beyer (D-VA), Del. James Moylan (R-GU)

Support

Advances generative-AI labeling and related consumer protections so people can tell when they are interacting with synthetic media.

Status: Ordered reported, 35–0

Protect: Yes
Innovate: Yes
Security: Yes
Accountable: Yes
Durable: Yes

Labeling is not a complete safety regime, but the public cannot exercise judgment if it cannot tell machine-generated content from human speech. Unanimous committee support is a model for how this work should look.

AI safetyTransparencyChild safety

Source

H.R. 7058 · House · March 26, 2026

Foreign Adversary AI Risk Assessment and Diplomacy Act

Rep. Michael Baumgartner (R-WA)

Support

Requires the State Department to assess risks from foreign adversaries using generative AI for malicious activity.

Status: Ordered reported, 45–0

Protect: Yes
Innovate: Yes
Security: Yes
Accountable: Yes
Durable: Yes

AI policy that ignores hostile-state use is not serious. Regular diplomatic and intelligence assessments are a minimum requirement for a country that intends to lead.

National securityCybersecurity

Source

H.R. 8819 · House · May 14, 2026

Federal Artificial Intelligence Risk Management Act of 2026

Rep. Ted Lieu (D-CA), Rep. Zach Nunn (R-IA), Rep. Don Beyer (D-VA)

Support

Requires federal agencies to use NIST’s AI Risk Management Framework when they deploy artificial intelligence.

Status: Introduced; referred to committee

Protect: Yes
Innovate: Yes
Security: Yes
Accountable: Yes
Durable: Yes

The federal government should not be an unaccountable early adopter. Putting NIST’s existing framework into agency practice is a practical, non-theatrical step.

Human accountabilityAI safety

Source

House (new) · House · September 3, 2026

Stop Rogue AI Act

Rep. Josh Gottheimer (D-NJ) and Rep. Mike Lawler (R-NY)

Support

Directs NIST to publish standards for securely deploying AI agents, including verification of actions and tamper-proof logs.

Status: Introduced September 2026

Protect: Yes
Innovate: Yes
Security: Yes
Accountable: Yes
Durable: Yes

Autonomous agents that take real-world actions without logs or owners are a governance failure waiting to happen. Standards that keep humans accountable for agent behavior should be table stakes.

CybersecurityAI safetyHuman accountability

Source

Discussion draft · Draft · June 4, 2026

Great American Artificial Intelligence Act

Rep. Jay Obernolte (R-CA) and Rep. Lori Trahan (D-MA)

Mixed

A wide federal AI framework including model safety and workforce provisions, with a three-year preemption of state laws that specifically regulate AI model development.

Status: Discussion draft released

Protect: No
Innovate: Yes
Security: Yes
Accountable: No
Durable: No

National standards can be wiser than a patchwork. A blunt three-year freeze on state development rules, including in child safety and biosecurity, may lock in gaps just as capabilities jump. Preemption should follow a real federal floor, not precede it.

American innovationWorkforce preparednessAI safety

Source

Senate (new) · Senate · September 4, 2026

Ban Artificial Superintelligence Act

Sen. Bernie Sanders (I-VT)

Oppose

Would permanently halt development of superintelligent AI and pause advanced AI until a new federal regulator sets guardrails, with criminal penalties for violations.

Status: Introduced September 2026

Protect: Yes
Innovate: No
Security: No
Accountable: Yes
Durable: No

Fear cannot be America’s AI strategy. A blanket halt cedes the frontier to competitors who will not pause, and criminalizing research is not a durable safety architecture. Serious evaluation, targeted high-risk controls, and human accountability are the better path.

AI safetyAmerican innovation

Source

Positions reflect American AI Safeguards’ five-question test applied to publicly described legislation. They are not endorsements of candidates. Legislator scorecards will be added when there are recorded votes to evaluate.