BenevolentAI vs Verge Genomics (2026)
Two approaches to finding a target in neurology, where almost everything fails in translation. BenevolentAI reasons over what has already been published, building a knowledge graph across literature, patents, genetics and chemistry, with the most externally adjudicated evidence record in this category and the only direct engagement with bias anyone here offers. Verge refuses the published corpus and the animal model alike, building on a proprietary library of multi omic human tissue, and it has taken a molecule into human study. The failure modes are mirror images. A knowledge graph faithfully reproduces the literature's blind spots; a tissue platform inherits the demographics of whoever donated the samples, and neither company publishes that composition.
- The knowledge graph reasons over the whole published corpus, linking genes, proteins, diseases and compounds from literature, patents, genetics and chemistry.
- The evidence record is the most externally adjudicated in this category, earning its standing on independence rather than volume.
- It addresses the bias question directly in its own words, which no other vendor located in this category does.
- The all in human platform is built on a proprietary library of multi omic patient tissue rather than animal or cell line proxies, addressing the translation failure that kills neurology programmes.
- Deal architecture is unusually well disclosed for a private company, so a partner can see the shape of the relationship before entering it.
- It has taken a molecule through investigational new drug status into human study, which most platforms here have not.
This comparison is published by AI Health Index, an independent research platform that compares healthcare AI vendors objectively. BenevolentAI and Verge Genomics are each graded against the same capability taxonomy, from each vendor's own public materials and the regulatory record, under the AI Health Index verification standard. No vendor pays for placement, and no vendor has reviewed this page. How this evidence is graded
Plain facts
| Fact | BenevolentAI | Verge Genomics |
|---|---|---|
| Primary category | Drug Discovery AI | Drug Discovery AI |
| Founded | Not recorded | 2015 |
| Headquarters | London, United Kingdom | South San Francisco, California, United States |
| Website | benevolent.com | vergegenomics.com |
Side by Side
Each record in one paragraph
Written to be quoted whole. Each paragraph states what the AI Health Index verified about the vendor, with the caveats attached. Generated from this pair’s live capability grades, so it moves when a grade moves.
The AI Health Index awards BenevolentAI its top capability grade on AI Centrality and Clinical and Operational Evidence. Set against Verge Genomics, BenevolentAI grades higher on several axes, including Model Supply Chain Disclosure, Clinical and Operational Evidence and AI Safety and PHI Stewardship. Its thinnest published disclosure sits on AI Liability and Recourse. Grades reflect evidence the AI Health Index could verify at the last review, so a low grade records disclosure the vendor has not published rather than a capability it has been shown to lack.
Source: AI Health Index, August 2026
The AI Health Index awards Verge Genomics its top capability grade on AI Centrality. Set against BenevolentAI, Verge Genomics grades higher on AI Liability and Recourse and Commercial Transparency. Its thinnest published disclosure sits on Model Supply Chain Disclosure. Grades reflect evidence the AI Health Index could verify at the last review, so a low grade records disclosure the vendor has not published rather than a capability it has been shown to lack.
Source: AI Health Index, August 2026
Questions buyers ask
Should we choose BenevolentAI or Verge Genomics?
On the axes where the AI Health Index separates them, BenevolentAI grades higher on several axes, including Model Supply Chain Disclosure, Clinical and Operational Evidence and AI Safety and PHI Stewardship, and Verge Genomics grades higher on AI Liability and Recourse and Commercial Transparency. BenevolentAI leads on the greater share of scored axes, but the split means the decision turns on which constraint is binding rather than on an overall winner.
Where do BenevolentAI and Verge Genomics differ most?
The widest separation the AI Health Index records between BenevolentAI and Verge Genomics is on Model Supply Chain Disclosure, where BenevolentAI grades C and Verge Genomics grades D. That axis sits in the AI Capability group, so it should carry the most weight for a buyer whose binding constraint is how much of the work the model itself is trusted to do.
Where do BenevolentAI and Verge Genomics grade the same?
The AI Health Index grades BenevolentAI and Verge Genomics the same on several axes, including AI Centrality, Autonomy and Oversight Model and Model and Technology Transparency. Neither holds an advantage the index can evidence on those axes, so they should not carry weight in a selection between these two.
What have BenevolentAI and Verge Genomics not disclosed?
At the last review, at least one of BenevolentAI and Verge Genomics published thin or absent detail on Model Supply Chain Disclosure and AI Liability and Recourse. The AI Health Index treats an absent disclosure as a gap in the public record rather than a failure of the product, so these are the axes to get in writing during diligence instead of inferring from the grade.
Related comparisons
Other published head to head assessments involving these vendors or their closest peers. The full set for this category is on the Drug Discovery AI page.
The failure modes are opposite and worth naming: a knowledge graph inherits the biases of the literature it was built from, so under studied diseases and populations stay under studied while the model looks confident, and a platform built on patient tissue inherits the demographic composition of whoever donated it, which neither vendor publishes. BenevolentAI's commercial disclosure has thinned since it left public markets, so figures from its listed period should not be assumed current. Neither holds a located security attestation.