Anthropic Partners with Six Medicine Developers as Claude Enters Drug Discovery

There’s a moment in any technology story where the abstract becomes very concrete. For Anthropic, that moment appears to have arrived in the life sciences.

The San Francisco-based AI company — best known for its Claude family of AI models — has announced it is working with six medicine-development organisations and has begun running drug-discovery programmes of its own. The announcement, posted by Anthropic on social media, describes a programme that spans everything from protein design to clinical-trial planning.

The six named partners are BigHat Biosciences, Bristol Myers Squibb, Divergence Labs, Enveda, Generate Biomedicines, and The Column Group. Anthropic says the collaborations have been running for several months, though exact start dates for each partnership aren’t specified.

What Claude Is Actually Doing in the Lab

The scope of the work, as Anthropic describes it, is broad. Claude is being used across multiple stages of drug development — scientific research, protein and antibody design, clinical-trial planning, and operational tasks such as reviewing manufacturing documentation.

One result stands out. Scientists at Generate Biomedicines used Claude Code to propose a redesigned CAR-T-cell signalling domain — a component of a type of cancer therapy that works by engineering a patient’s own immune cells to attack tumours. Anthropic says the proposed design was about 70% shorter than the standard version. In laboratory tests, the redesigned domain reportedly enabled immune cells to produce roughly ten times as many receptors.

That’s a striking early-stage finding. But it’s worth being clear about what it is and what it isn’t. These are company-reported laboratory results, not peer-reviewed published findings. No medicine arising from this work has received regulatory approval, and the figures haven’t been independently verified. Early lab results in drug development frequently don’t survive the long journey through clinical trials and regulatory scrutiny.

Bristol Myers Squibb, one of the world’s largest pharmaceutical companies, is using Claude differently. According to Anthropic, BMS is applying the AI to assess early-stage clinical-trial plans and estimate the likelihood of trial success. That kind of predictive work — helping researchers decide which trials are worth pursuing — could, if it proves reliable, reduce the enormous cost and time involved in drug development.

Anthropic’s Own Drug Programmes

Beyond the partnerships, Anthropic says it has started internal drug-discovery efforts. One area of focus is rare childhood forms of muscular dystrophy — conditions that currently have very limited treatment options. The company hasn’t specified how many programmes are under way or at what stage they sit.

Anthropic has also launched what it calls Claude Science, described as an AI workbench aimed at researchers, and says it is providing Claude credits for life scientists.

Anthropic says the bottlenecks are “distributed all throughout the drug development process”, and that its strategy is to have AI increasingly involved in all of those tasks.

A Growing Race in AI-Assisted Drug Discovery

Anthropic isn’t alone in this space. Google DeepMind’s AlphaFold has already reshaped how researchers approach protein structure prediction. OpenAI has explored biomedical applications. And a string of dedicated biotech AI companies have raised hefty funding on the promise of faster drug discovery.

But Anthropic’s move into active partnerships with named pharmaceutical companies — including a firm of Bristol Myers Squibb’s scale — signals that the major AI model developers are moving beyond general-purpose tools and into domain-specific applications.

The regulatory picture remains complex. In the UK, any medicine developed with AI assistance would still need to go through the Medicines and Healthcare products Regulatory Agency’s approval process. The MHRA has been developing guidance on AI in healthcare, but no regulatory framework specifically covering AI-designed medicines is yet in place. And questions about accountability — who is responsible if an AI-assisted design causes harm — remain genuinely open.

What the Sceptics Are Saying

Not everyone is convinced the hype matches the reality. The history of drug development is littered with promising early results that didn’t translate into effective treatments for patients. AI tools can accelerate certain steps in the research pipeline, but they can’t bypass the fundamental biology, the need for clinical evidence, or the regulatory requirements that protect patients.

There are also questions about data. Training AI models on proprietary pharmaceutical data raises issues around confidentiality, intellectual property, and the potential for bias if training datasets don’t reflect diverse patient populations.

And the performance figures Anthropic is reporting — ten times more receptors, a 70% shorter design — come from the companies themselves. Independent replication, peer review, and longer-term clinical evidence are what the scientific community will need before these results change how medicine is actually made.

What This Means for Kent Residents

There’s no direct connection between Anthropic’s announced partnerships and any current NHS service, clinical trial, or research programme in Kent. However, pharmaceutical and biotech research is active across the South East, and any AI tools that genuinely speed up the drug-discovery pipeline could eventually affect which medicines reach NHS patients — including those treated by NHS Kent and Medway. For now, the practical impact on Kent residents is distant; this is early-stage research, and any medicine emerging from it would need full MHRA approval before reaching a prescription.