About TrustHawk
TrustHawk is an independent benchmark that rates AI companies, not their models, on safety, transparency, security and privacy using verified public data.
Why TrustHawk exists
More of our work, conversations and personal data now pass through AI products. Model benchmarks tell you how capable a model is. They don't tell you whether the company behind it is careful, open and trustworthy with your data. TrustHawk answers that second question.
Other AI rankings
Rank how capable a model is
- Accuracy
- Speed
- Reasoning
TrustHawk
Rates whether the company is careful
- Safety
- Transparency
- Security
- Privacy
Why we built this
TrustHawk exists to hold AI companies accountable for the choices they make on safety, transparency, security and privacy. By scoring every company the same way, in public, we want to encourage them to compete to be the best on all four. When they do, everyone benefits.
Companies, not models
A company can ship many models, and its safety work, security record and privacy policy cover all of them. So TrustHawk rates the company as a whole, on four equal pillars: Safety, Transparency, Security and Privacy.
One rating
The company
That rating covers every model and product the company ships.
Independent and evidence based
No payment
No company pays to be rated, and no company can pay to change its score.
Same method
Every score comes from verified public data, and every company is rated the same way.
Public sources
Every company page lists the public sources behind its score, so anyone can check them.
No guessing
When an outside rating does not cover a company, the score uses the average of the companies it covers. We do not guess an answer for a check we cannot confirm.