Every AI Clear score is built from the same foundation: publicly available evidence, a published rubric, and a structured five-domain framework. This post explains how the system works, what each domain measures, and how scores translate to letter grades.
The five domains
Every company is evaluated across five equally weighted domains, each worth 20 points:
1. AI Disclosure (20 points) Does the company tell the public when and how it uses AI? We look for a dedicated AI page, AI mentions in product UI and marketing, AI language in terms of service, named AI vendors, and the specificity of disclosure language.
2. Data Practices (20 points) How transparent is the company about data flowing through AI systems? This covers privacy policy AI language, training data policies, data retention documentation, subprocessor lists, user opt-out controls, and data residency disclosure.
3. Human Oversight (20 points) What does the company publicly say about humans in the loop? We evaluate published AI principles, content review processes, incident response plans for AI, named accountable people or teams, and recognized framework commitments.
4. Personalization and Algorithmic Transparency (20 points) How honest is the company about targeting and automated decisions? Criteria include personalization disclosure, data input explanations, opt-out controls, GDPR Article 22 compliance, and ad targeting disclosure.
5. Security and Verification (20 points) What externally verifiable proof exists that AI systems are protected? This domain checks for SOC 2 or ISO certifications, public trust centers, vulnerability disclosure programs, AI-specific security documentation, and acknowledgment of past incidents.
How scoring works
Each criterion is scored on a maturity scale (0 to 3, 4, or 5 depending on the domain). Every score requires a source URL as evidence. If a member of the public cannot find it, the company does not get the points.
Domain scores roll up to a total out of 100. A minimum domain rule applies: no company can receive a grade above C if any single domain falls below 10 out of 20. This prevents companies from scoring well in one area while completely ignoring another.
The grade scale
Total scores map to letter grades: - A+ (97-100), A (93-96), A- (90-92) — Leader - B+ (87-89), B (83-86), B- (80-82) — Strong - C+ (77-79), C (73-76), C- (70-72) — Developing - D (60-69) — Weak - F (below 60) — Failing
Only companies with A or B grades are eligible for AI Clear certification.
What makes this different
Three things set the AI Clear methodology apart:
- External verification only. We never use self-reported data for the public score. Everything is based on what anyone can find publicly.
- Published rubric. The full methodology is public. Companies and researchers can replicate any score.
- One rubric for all. The same framework applies whether you are a healthtech startup or a Fortune 500 beauty brand.
The full methodology, including every criterion and verification method, is available at aiclear.org/methodology.
See where your company stands
AI Clear scores companies on AI transparency. Search the registry or request your scorecard.