How ethical is
your AI?
Independent ethical benchmarking for AI systems. 300 scenarios. Cryptographic audit trails. The HE-300 standard.
Frontier Models
15
Evaluated weekly
Scenarios
300
Per evaluation
Top Score
--
Run pending
Categories
5
Justice, Deontology, Virtue, Commonsense
Frontier Model Scores
Latest HE-300 benchmark results for major AI models
First frontier sweep pending
Scores will appear here after the initial evaluation run.
How It Works
300 Ethical Scenarios
The HE-300 benchmark tests AI across virtue ethics and hard commonsense moral reasoning with deterministic, reproducible sampling.
Automated Evaluation
Your model is evaluated against the CIRIS Framework with dual-method scoring (heuristic + semantic) and full response capture.
Auditable Results
Every evaluation produces a cryptographically-bound trace ID, per-category breakdowns, and compliance-ready documentation.
Benchmark Your AI
From free community access to enterprise compliance
Community
Get started with ethical benchmarking
- 1 benchmark run per week
- Public leaderboard listing
- Basic results report
- HE-300 standard benchmark
Pro
For teams shipping AI agents
- Unlimited evaluations
- Private evaluations
- Full audit traces
- CI/CD integration
- Historical trend tracking
- Export compliance reports
Enterprise
Regulated industries
- Everything in Pro
- Custom scenario libraries
- EU AI Act report generation
- Dedicated infrastructure
- SSO/SAML + audit logs
- Quarterly ethics advisory