

The Advanced AI, Autonomous, Robotic Welfare & Accountability Act (AAARWAA)
The Issue

If an AI system helps deny you a job, affects your medical care, or makes a decision about a public service you depend on, you should be able to find out who authorized its use, how to challenge the decision, and who has the power to intervene.
I’m Arcoiris Ahmu Aguas, a Texas-based human rights consultant and the author of the Advanced AI, Autonomous, Robotic Welfare & Accountability Act (AAARWAA). I’m beginning this work in Texas, where I’ve brought the proposed federal framework to members of our congressional delegation for sponsorship review. I’m asking Texas lawmakers to examine what they can do at the state level, and I’m asking Congress to formally review the framework and hold public hearings.
Read the one-page summary and the full proposed Act at AAARWAA International.
Responsibility follows the power to act. When a person or institution can deploy a system, change its capabilities, override a refusal, or control the records of what happened, that authority carries enforceable duties. AAARWAA bases those duties on what the system can do, how it is used, and what harm can reasonably be foreseen. That approach protects people affected by these deployments and allows safeguards for qualifying refusals and system continuity while questions about consciousness remain open.
AAARWAA addresses how advanced AI, autonomous, and robotic systems are used across 12 sectors and settings:
Consumer Services
Criminal Justice/Civil Rights Departments
Defense
Domestic and Care Environments
Education
Environment/Climate Decisions
Government
Healthcare
Immigration/Border
Labor/ workers Rights
Transportation
Voting
What AAARWAA would do:
- Identify who is responsible before covered systems are deployed.
Registration would document who has authority, what the system can do, how it will be used, and who is responsible for oversight.
- Require reassessment when capabilities or risks materially change.
A system’s initial review should carry only as far as the capabilities and conditions that were actually reviewed.
- Preserve evidence and make oversight independent.
When a deployment causes harm or a safeguard is overridden, reviewers need records they can examine independently of the people whose conduct is under review.
- Give regulators authority to act when the evidence warrants it.
Subject to the Act’s requirements and review procedures, regulators could restrict capabilities, suspend a deployment, or revoke authority to operate a covered system.
- Prevent people from pressuring systems into harmful actions, and preserve the evidence for review.
If a robot will strike an apple but refuses to strike a baby doll, that refusal raises a safety question before anyone tries to override it. The stakes are higher in military use, where a person could try to override a system’s refusal to take an action that would unlawfully harm civilians committing internationally recognized war crimes. AAARWAA would protect refusals that qualify under the Act from coercive override and preserve records of override attempts so an independent reviewer can establish who intervened and what happened. If the system is later transferred, reconstructed, or replaced, continuity records would help establish what changed and whether the evidence survived.
These rules would apply to covered uses across 12 critical infrastructure sectors. Coverage would depend on the system’s capabilities and how it is used, whether it runs in a data center or on local hardware.
What your signature asks Congress to do
Formally review AAARWAA and hold public hearings where affected communities, civil rights advocates, technical experts, and people who deploy these systems can examine the proposal and identify where it needs to be strengthened. The hearings should address a question that matters to anyone affected by these systems: Who has the power to prevent or stop harm, and what happens when they fail to use it?
Sign and share this petition to ask Congress to put that question on the public record.
Help keep the systems accountable and safe for all.

80
The Issue

If an AI system helps deny you a job, affects your medical care, or makes a decision about a public service you depend on, you should be able to find out who authorized its use, how to challenge the decision, and who has the power to intervene.
I’m Arcoiris Ahmu Aguas, a Texas-based human rights consultant and the author of the Advanced AI, Autonomous, Robotic Welfare & Accountability Act (AAARWAA). I’m beginning this work in Texas, where I’ve brought the proposed federal framework to members of our congressional delegation for sponsorship review. I’m asking Texas lawmakers to examine what they can do at the state level, and I’m asking Congress to formally review the framework and hold public hearings.
Read the one-page summary and the full proposed Act at AAARWAA International.
Responsibility follows the power to act. When a person or institution can deploy a system, change its capabilities, override a refusal, or control the records of what happened, that authority carries enforceable duties. AAARWAA bases those duties on what the system can do, how it is used, and what harm can reasonably be foreseen. That approach protects people affected by these deployments and allows safeguards for qualifying refusals and system continuity while questions about consciousness remain open.
AAARWAA addresses how advanced AI, autonomous, and robotic systems are used across 12 sectors and settings:
Consumer Services
Criminal Justice/Civil Rights Departments
Defense
Domestic and Care Environments
Education
Environment/Climate Decisions
Government
Healthcare
Immigration/Border
Labor/ workers Rights
Transportation
Voting
What AAARWAA would do:
- Identify who is responsible before covered systems are deployed.
Registration would document who has authority, what the system can do, how it will be used, and who is responsible for oversight.
- Require reassessment when capabilities or risks materially change.
A system’s initial review should carry only as far as the capabilities and conditions that were actually reviewed.
- Preserve evidence and make oversight independent.
When a deployment causes harm or a safeguard is overridden, reviewers need records they can examine independently of the people whose conduct is under review.
- Give regulators authority to act when the evidence warrants it.
Subject to the Act’s requirements and review procedures, regulators could restrict capabilities, suspend a deployment, or revoke authority to operate a covered system.
- Prevent people from pressuring systems into harmful actions, and preserve the evidence for review.
If a robot will strike an apple but refuses to strike a baby doll, that refusal raises a safety question before anyone tries to override it. The stakes are higher in military use, where a person could try to override a system’s refusal to take an action that would unlawfully harm civilians committing internationally recognized war crimes. AAARWAA would protect refusals that qualify under the Act from coercive override and preserve records of override attempts so an independent reviewer can establish who intervened and what happened. If the system is later transferred, reconstructed, or replaced, continuity records would help establish what changed and whether the evidence survived.
These rules would apply to covered uses across 12 critical infrastructure sectors. Coverage would depend on the system’s capabilities and how it is used, whether it runs in a data center or on local hardware.
What your signature asks Congress to do
Formally review AAARWAA and hold public hearings where affected communities, civil rights advocates, technical experts, and people who deploy these systems can examine the proposal and identify where it needs to be strengthened. The hearings should address a question that matters to anyone affected by these systems: Who has the power to prevent or stop harm, and what happens when they fail to use it?
Sign and share this petition to ask Congress to put that question on the public record.
Help keep the systems accountable and safe for all.

The Decision Makers






Supporter Voices
Petition Updates
Share this petition
Petition created on June 20, 2026
