we_are_coded.by CODE · The world, decoded
БГ

Anthropic · Oversight and access

Anthropic opened Claude to verified biologists with fewer safeguards, and for risky projects without the biology ones

The life sciences verification programme gives approved teams more permissive classifiers. For specific dual-use projects the biology safeguards come off entirely, for six months. The price is 30 days of data retention for monitoring.

Source
Anthropic
Published
The event
Checked on
Responsible editor
Cvetelin Ivanov
How we work
MethodCorrections

A virologist making a vaccine and someone who wants a virus to spread more easily ask almost the same questions. The classifier often cannot tell them apart by the words, so until now it stopped both.

Anthropic changes the question. Instead of “what are you asking” it checks “who are you”.

Fact

The facts: on 17 September 2026 Anthropic introduced the Life Sciences Verification Program (LSVP) in beta, initially for teams and institutions. Applicants go through a review of their research credentials, security standards and ethical oversight. Those approved get two types of access. Standard Use is for the whole team, is renewed once a year and gives the Mythos 5.1, Opus 5 and Sonnet 5 models with classifiers more permissive for science than the generally available ones. High-risk Use is an add-on for one specific project, is renewed every six months and removes all safeguards that block life sciences requests; for Opus 5 and Sonnet 5 it is available now, while for Mythos it is for a small set of organisations with additional vetting for the time being, as the company works with the US government. The cyber classifiers remain. Monitoring shifts from real-time blocking to offline analysis of behaviour, which is why traffic is retained for 30 days; per Anthropic, this data is compartmentalised, not used for training and not accessible to its biology research teams. Dozens of organisations have already gone through early access, and the company expects to enrol hundreds in the first week.

What actually changes

Until now the safeguard stood at every door and questioned every request. Now it lets the verified in and reviews the record afterwards, to catch a pattern spread across many sessions.

A pass at the entrance and a log inside.

The risks are named by them: a hijacked account, an employee working for someone else, and an agent that does something dangerous during a long task without anyone asking it to. The third is the newest. Vetting the person does not vet the agent they leave running overnight.

And one small but important detail - the 30 days of retention. For a lab that is a small price. For a company guarding someone else's intellectual property, it is a line in the contract that some lawyer will want to read twice.

If you work in a lab that has been getting refusals for normal work, the path is through the waitlist. If you do not, the news for you is different: the boundary is no longer only in the model but also in the vetting of the people around it.

The visual is generated code-art, no third-party images.