Artificial intelligence firm Anthropic has launched enhanced capabilities for its Claude system aimed at supporting scientific research workflows, from initial discovery through commercial translation.

The San Francisco-based company, which positions scientific progress as central to its public benefit mission, announced improvements designed to make Claude more useful for life sciences professionals including researchers, clinical coordinators and regulatory affairs managers.

Anthropic stated that whilst scientists previously employed Claude for discrete tasks such as coding statistical analyses or summarising papers, the updated system aims to support entire research processes rather than isolated components.

The company's latest model, Claude Sonnet 4.5, demonstrates improved performance on life sciences benchmarks. On Protocol QA—which tests understanding of laboratory protocols—Sonnet 4.5 scores 0.83 against a human baseline of 0.79 and its predecessor's 0.74 performance. Similar improvements appear on BixBench, an evaluation measuring bioinformatics task performance.

Anthropic is introducing connectors allowing Claude to access scientific platforms directly, including:

• Benchling, enabling responses with links to source experiments, notebooks and records
• BioRender, providing access to scientific figures, icons and templates
• PubMed, offering millions of biomedical research articles and clinical studies
• Scholar Gateway from Wiley, supplying peer-reviewed scientific content
• Synapse.org, allowing data sharing and analysis in public or private projects
• 10x Genomics, enabling single cell and spatial analysis through natural language

These scientific connectors supplement existing integrations with general-purpose tools including Google Workspace, Microsoft applications, Databricks for large-scale bioinformatics analytics, and Snowflake for natural language dataset queries.

The company has also released Agent Skills—folders containing instructions, scripts and resources that Claude uses to perform specific tasks consistently. Anthropic describes these as particularly suited to scientific work requiring adherence to specific protocols and procedures.

The firm is developing scientific skills beginning with single-cell-rna-qc, which performs quality control and filtering on single-cell RNA sequencing data using established best practices. Scientists can also create custom skills for their requirements.

Anthropic suggests Claude can support research tasks including literature reviews and hypothesis development, protocol generation, bioinformatics and data analysis, and clinical and regulatory compliance. The company is creating a prompt library designed to elicit optimal results for such applications.

To support adoption, Anthropic is providing hands-on assistance from subject matter experts whilst partnering with consultancies including Caylent, Deloitte, Accenture, KPMG, PwC, Quantium, Slalom, Tribe AI and Turing, alongside cloud partners Amazon Web Services and Google Cloud.

The company noted that existing customers and partners have already employed Claude for various real-world scientific tasks, though it provided limited specifics.

Beyond commercial offerings, Anthropic supports life sciences research through its AI for Science programme, which provides complimentary API credits to researchers working on high-impact scientific projects globally. The company stated these partnerships help identify new Claude applications whilst assisting scientists with pressing questions.

Claude for Life Sciences is available through Claude.com and Amazon Web Services Marketplace, with Google Cloud Marketplace availability forthcoming.

The announcement reflects broader industry efforts to position AI systems as scientific research tools. Whether such systems genuinely accelerate discovery or merely automate routine tasks remains debated, with some researchers expressing enthusiasm whilst others question whether current AI capabilities truly advance scientific understanding beyond incremental efficiency gains.

Anthropic's focus on life sciences specifically targets an industry increasingly exploring AI applications across drug discovery, clinical trials, regulatory submissions and data analysis. Pharmaceutical and biotechnology companies represent potentially lucrative customers willing to pay premium rates for tools promising to accelerate research timelines or reduce development costs.

However, scientific AI applications face particular scrutiny regarding accuracy and reliability. Mistakes in literature summarisation, protocol generation, or data analysis could have serious consequences, raising questions about appropriate human oversight and validation of AI-generated scientific work.

The company's emphasis on connecting Claude to established scientific platforms rather than replacing them suggests recognition that integration with existing workflows proves more practical than attempting wholesale replacement of current research infrastructure

To learn more about Claude for Life Sciences or set up a demo with our team, see here.

Claude for Life Sciences is available through Claude.com and on the AWS Marketplace, with Google Cloud Marketplace availability coming soon.