Etika Policy Institute
Making AI
accountable.
The Etika Policy Institute is a working idea institute. Directors, scholars, former regulators, and operators publish concrete memos on how artificial intelligence should be governed, audited, and deployed.
Each year, the strongest proposals are adopted into the next revision of the AIMSS standard. The standard is not static — it evolves with the best ideas the field can offer.
2025 Cycle
- Memos published
- 4
- Adopted into AIMSS v1.1
- 0
- Under review
- 0
- Open for response
- 4
Submit a memo
The Institute accepts memos by invitation and through an open call each spring. Drafts must include a problem statement, a concrete proposal, and a section on implementation costs and risks.
institute@etika.foundation →How proposals become standard
01
Memo published
Author releases a memo on this site. Open public comment for 60 days.
02
Institute review
A standing committee of directors and external reviewers grades for evidence and feasibility.
03
Pilot in audits
Promising proposals are tested as voluntary clauses in real Etika audits over 6–12 months.
04
Adoption vote
The Board votes annually on which clauses become mandatory in the next AIMSS revision.
Governance
2025 · 12 min
Ethical Governance and Human Accountability
Mandatory AI Ethics Governance Framework approved by senior leadership: prohibited uses, ethical review triggers, named business/technical/risk owners, an independent Ethics Committee, AI Ethical Impact Assessments for high-risk systems, user disclosures, and a non-retaliatory escalation channel. Failures classified as major or critical nonconformity.
Aiden Muscovitch · Etika Policy Institute
Open for response
Safety
2025 · 11 min
AI Security, Resilience, and Misuse Prevention
AI-specific security controls across the lifecycle: threat assessments, red-team testing for high-risk systems, model-integrity and access controls, abuse detection and rate limits, supplier security review, fallback and shutdown procedures, and severity-classified incident reporting to the certification body.
Aiden Muscovitch · Etika Policy Institute
Open for response
Climate
2025 · 10 min
Environmental Sustainability and Responsible Compute
An AI Environmental Impact Policy covering material training, fine-tuning, and inference: documented compute and emissions estimates, responsible-compute review at release approval, ongoing monitoring of energy and accelerator usage, hardware lifecycle accountability, and proportionate public reporting that distinguishes measured data from estimates.
Aiden Muscovitch · Etika Policy Institute
Open for response
Labour
2025 · 12 min
Workplace Safety, Labour Protection, and Social Responsibility
Labour and social-responsibility requirements covering employees, contractors, annotators, moderators, evaluators, and material suppliers: psychological safety controls for harmful-content work, fair and transparent compensation, supplier due diligence, community and accessibility impact assessment, and confidential grievance and whistleblower channels.
Aiden Muscovitch · Etika Policy Institute
Open for response
A standard that learns
The AIMSS standard is revised every year, in public.
Read the standard, see which Institute proposals shaped it, and watch the next revision form in real time.