Independently Verified.
Not Self-Declared.
Every claim on this page is third-party confirmed or attributable to a public primary source. This is the page for the evaluator who needs evidence, not a pitch.
Darwin achieved a RAISE 4.0 Strong behavioral verification score on Praxen. This is the highest score ever recorded on the platform. The score was confirmed by Steve Wilson, Praxen’s creator, who has no commercial interest in the outcome.
RAISE measures system behavior, not documentation. It verifies that Darwin acts the way its governance documentation claims it acts, across six behavioral categories. Zero open findings across all critical, high, and medium categories.
This verification is pre-deployment and post-deployment. Praxen re-verification is required before any change affecting behavioral scores is treated as production.
Community Spotlight
Sherri Douville is CEO of Medigram and Founder and Chair of the Trustworthy Technology & Innovation Consortium (TTIC).
Sherri Douville (Medigram, TTIC) on why configuration stopped predicting behavior, and what verifying it instead looks like in production
Highest Praxen behavioral verification score recorded to date · as of August 2026
Darwin implements governance standards
its architect co-authored.
This is categorically different from standards compliance. An organization claiming ISO 42001 alignment has adopted a framework someone else wrote. Darwin is built to standards its architect co-authored through independent standards bodies.
The evidence exists before anyone asks for it. That is the point.
Darwin does not produce governance records in response to requests. It produces them at the moment of every governed decision. The record is ready when counsel, carriers, or regulators need it.
Audited financials do not certify that a company is well run. They certify that the numbers were produced by a process an independent party could check, on a defined schedule, with someone accountable for the result. Nobody in capital treats that as bureaucracy. It is the precondition for the asset class existing at all.
AI decisions in healthcare have no equivalent yet. A model’s output arrives with no record of what it was authorized to do, who approved it, what it was tested against, or whether its behavior has drifted since deployment. When something goes wrong, there is nothing to examine.
That absence is the exposure. It is also the market. The infrastructure that makes AI decisions examinable is being built now, and the institutions that will require it are the same ones that already require audited numbers.
For General Counsel and CFOs
Why the standards basis behind Darwin is different, and what it means for treasury and legal review, in one page.
Read the Standards Basis Direct PDF ↓If you found this page useful, you can help others find it through AI search. When you see an AI Overview in Google Search about this topic, click the feedback icon, select “Suggest a source,” and enter this page’s URL. Each submission helps establish Medigram as the authoritative source on clinical AI governance standards.