← Back to Home

Novo Partners with Anthropic to Take Claude from Trial Documentation into Drug Discovery

The partnership expands AI’s role from accelerating document production to helping understand biology and drug action. The initial R&D workflows are still in the test-planning stage, and whether they can improve the quality of drug candidates remains publicly unverified.

By SURL BioNews

When a pharmaceutical company’s AI tools can quickly complete trial documents, they save administrative time; when they help assess biological mechanisms and identify directions for new drugs, the challenge extends into research decision-making. Danish drugmaker Novo and Anthropic announced a partnership on September 16, 2026, to test Claude Science in selected R&D workflows, extending existing document automation applications into drug discovery.

According to Novo’s announcement, the company’s scientists and computational teams will identify key questions in drug discovery, after which the two companies will develop corresponding tools and workflows to support biological reasoning. This means the drugmaker’s teams will continue to define the research questions, with initial testing focused on steps both companies consider most likely to deliver benefits. The announcement did not identify specific diseases, molecular targets, or drug candidates.

A more concrete existing use comes from Reuters’ reporting: Novo has already used Claude to automatically generate trial reports and reduce the time needed to produce patient-related documents from months to minutes. However, the report did not specify the document types, comparison methods, or time required for human review. This description of efficiency also cannot be directly translated into shorter clinical trials or overall drug development timelines.

The new partnership has not yet released data that could be used to measure scientific outcomes. The companies have not disclosed test datasets, benchmarks for comparing model performance, or experimentally validated research findings. What can currently be confirmed is therefore the direction of the partnership and its initial deployment arrangements. It is not yet possible to determine whether Claude can improve the accuracy of biological hypotheses or help researchers select drug candidates with a greater chance of success.

The partnership also includes using Anthropic models to support software development, laying the tooling foundation for Novo to expand its use of AI. BioPharma Dive noted that the companies did not disclose financial terms. The available information is insufficient to estimate the scale of the partnership, and no specific arrangements concerning drug rights are available for assessment.

### Background Context

As AI tools move from the research workbench into major pharmaceutical companies, the questions they face become more concrete: how research teams verify outputs, trace data, and decide which inferences warrant experimental investment. Novo said the partnership incorporates data governance and human oversight and must meet the company’s ethical and compliance standards. However, it has not publicly detailed the scope of data access, review procedures, or error handling.

These arrangements will influence how far the tools can go. Document production speed can be measured in time, while the value of biological reasoning must be assessed through reproducible analyses and experimental results. If the tools are later used to support clinical development or regulatory submissions, demonstrating the reliability of data and analyses will also become necessary. This announcement did not provide validation evidence at that level.

References

  1. Euronews.com
  2. Novo Nordisk
  3. Reuters via Euronext