Anthropic opens a verification programme for life science users
Verified professionals gain access to a wider range of biology work, including a model the company has not previously opened to it.
Anthropic has opened applications for what it calls a Life Sciences Verification Programme, under which professionals working in the life sciences can use its models with a different set of safeguards from the default.
The company said the programme gives access to models including Mythos, which it had not previously made available for this category of work. The stated aim is to let verified users carry out the full range of biology-related tasks.
The design acknowledges a problem the company has written about at length. Safeguards tuned to prevent misuse of biological capability also obstruct legitimate researchers, and a blanket restriction pushes real work elsewhere rather than making it safer.
Verification shifts that tradeoff rather than resolving it. Access now depends on Anthropic's assessment of who an applicant is, which concentrates the judgement inside the company rather than in the model's refusals.
The announcement is Anthropic's own. It has not published how applicants are assessed, how many have been approved, or what happens when a verified account is misused.