The independent trust and validation layer for technology
Trust in technology should be verified — not assumed.
Sustain AI Index independently validates technology companies and AI systems across Trustable AI, Security and Sustainability using evidence-based assessment and third-party verification.
Evidence-based. Auditor-verified. Published by choice.
The framework
One framework behind every published rating.
The full framework: 100 questions across 21 control areas, in three weighted pillars.
Technology claims are easy. Verification is harder.
Technology companies increasingly make claims around:
- Trustable AI
- Security
- Privacy
- Sustainability
- Compliance
- Ethical technology
Sustain AI Index addresses this gap.
- Each answer in the assessment describes a control the organisation has in place, and the auditor checks it against the evidence behind it. The published result shows the overall score, each pillar's score and how many mandatory findings were recorded.
- Every answer carries a supporting document or technical record. The auditor reviews it; it is never published. The registry shows what was verified, not the documents themselves.
- An independent third-party auditor reviews every answer and signs the result, and the organisation countersigns it before it can be published.
- The same 100 questions across 21 control areas for every limited- and high-risk system, rolled up into Trustable AI, Security and Sustainability. Each result records the scoring-engine version that produced it. Read the methodology.
- Each published result shows its assessment cycle and the date it was published.
- A result is valid for one year from its final signature. The organisation can withdraw it from the registry at any time, and it leaves public view immediately.
Three trust pillars
One weighted assessment across three dimensions.
Trustable AI
65% of the assessment · 66 questions
- Governance & accountability
- Risk management system
- Data governance & quality
- Technical documentation
- Transparency & explainability
- Human oversight
- Accuracy, robustness & reliability
- Fairness, bias & non-discrimination
- Conformity assessment & certification
- Ethical AI & fundamental rights
- Staff competence & AI literacy
- Inclusive design & accessibility
- Contestability & redress
- Maturity & continuous improvement
Security
25% of the assessment · 24 questions
- Cybersecurity
- Logging, monitoring & post-market surveillance
- Privacy & data protection
- Security operations
- Change management & lifecycle
Sustainability
10% of the assessment · 10 questions
- Sustainability & environmental responsibility
- Procurement & third-party obligations
The validation process
From documentation to verified trust.
- STEP 1
Company Profile
The supplier describes company, product, AI systems, purpose, users and workflow.
- STEP 2
Evidence Submission
Architecture, policies, documentation, risk assessments, test results and more.
- STEP 3
Structured Assessment
Evaluated against defined metrics and questions across the three pillars.
- STEP 4
Independent Auditor Verification
An independent auditor reviews responses, evidence and risk areas.
- STEP 5
Verified Public Result
The company can choose to publish the verified result in the public registry.
Independence
We don’t simply publish what companies say about themselves.
A supplier submits information and evidence. The assessment is then subject to independent review by an independent auditor. Every published result clearly separates what was claimed from what was verified.
Payment provides access to the assessment process — it does not determine the assessment outcome.
Every published result distinguishes
- Supplier-provided information
- Evidence submitted
- Assessment result
- Auditor verification
- Publication status
- Assessment date
- Validity period
- Methodology version
Differentiation
More than compliance. More than a rating.
Compliance tools
Help companies manage requirements.
Rating / review platforms
Provide opinions or market signals.
Sustain AI Index
Evidence + structured assessment + independent verification + public trust profile.