Solar System Union supplied sealThe Solar System UnionPublic institutions
Shared rights across worlds
My local casesDocumentsAccessibility
AI & Autonomous Systems Authority identity
AIR / authority
AI & Autonomous Systems Authority
UnionAI & Autonomous Systems AuthorityAI & Autonomous Systems Authority

AIR / PUBLIC SERVICE NETWORK

AI & Autonomous Systems Authority

Certifies high-impact autonomous systems and preserves meaningful human accountability.

AIRauthorityUnion-wide competence

Independent oversight of consequential AI

The AI and Autonomous Systems Authority independently examines consequential automated systems used in public services and shared infrastructure. It serves affected people, accountable operators, research teams and institutions procuring or deploying such systems. A technically impressive model does not acquire public authority by being installed.

Choose your route

Office output: Model assessment

Evidence, affected decisions and human remedy

Teams review safety evidence, training provenance, rights impacts, incident reports, robotic labour interfaces and the public algorithm register. Assessments distinguish assistive recommendations from systems that affect eligibility, access or operational permission. The record states the deployment purpose, model version, responsible human role and known limitations.

The powers we hold and the powers we refer

The Authority does not own personal data rights; Data Rights Authority retains that competence. A public operator must provide a meaningful human review route for consequential decisions, not only an explanation of the model's output. Independent testing and accessible challenge are required within the published assessment scope.

A case in practice

Illustrative case / AIR

A habitat utility uses a model to prioritise reserve access. A resident challenges an exclusion. The Authority reviews the specific deployment and impact, while the operator preserves essential continuity and a human decision-maker reassesses the case. A favourable laboratory benchmark alone cannot justify the disputed public restriction.

Responsibilities and programmes

Safety Evaluation

Safety Evaluation brings together model evaluation, autonomous craft assurance and life-support ai review.

Explore this programme →

Rights & Accountability

Rights & Accountability brings together decision explanation, appeal pathways and human oversight standards.

Explore this programme →

AI Incident Response

AI Incident Response brings together failure investigation, model recall coordination and evidence retention.

Explore this programme →

Research Governance

Research Governance brings together controlled evaluation environments and high-risk research review.

Explore this programme →

Public Algorithm Register

Public Algorithm Register brings together government model inventory and audit publication.

Explore this programme →

Where our powers stop

Certification is independent of vendor procurement and operational command.

Reporting line
The Solar System Union
Independent review
Union Ombudsman / Union Court of Justice

Public records and accountability

AIR/REG
Records and reasoned determinations

Annual service plan; public decisions; performance and expenditure report.

AIR/CONTACT
Contact the competent office

Accessible assistance, evidence alternatives, institutional handoffs and review.