Novo, the Danish pharmaceutical giant behind Ozempic, has partnered with Anthropic to use advanced AI models for drug discovery, marking the latest move by major drugmakers to integrate artificial intelligence across research and development pipelines. The collaboration signals a strategic shift in how the industry approaches the notoriously slow and expensive process of bringing new medicines to market.

Why Is Pharma Suddenly Doubling Down on AI Partnerships?

The partnership between Novo and Anthropic reflects a broader industry recognition that AI can fundamentally reshape how drugs are discovered. Rather than relying on a single AI vendor, Novo is building a portfolio approach, combining Anthropic's Claude models with other technology partners to maximize the impact of artificial intelligence across its research organization.

"Joining forces with Anthropic will supercharge our R&D organisation and help us on our mission to bring new, transformative health solutions to people living with chronic diseases," said Mike Doustdar, president and CEO of Novo.

Mike Doustdar, President and CEO of Novo

Novo will deploy Anthropic's Claude Science tool to advance scientific reasoning in research and development workflows. The company will test Claude on specific R&D challenges identified by its scientists and computational teams, with a focus on areas where the combined capabilities of Novo and Anthropic are expected to have the greatest impact.

The potential payoff is enormous. Dario Amodei, co-founder and CEO of Anthropic, framed the opportunity in striking terms, noting that AI's advancing capabilities could compress "a century's worth of biological and medical breakthroughs into a decade". This isn't hyperbole in the context of drug development, where traditional timelines stretch 10 to 15 years from initial discovery to patient access.

How Does Novo's AI Strategy Fit Into Its Broader Tech Ecosystem?

Novo is not putting all its eggs in one basket. The company already uses OpenAI's science-focused GPT model, Rosalind, after signing a deal in April 2026 to analyze complex datasets and identify potential drug candidates. Now, with Anthropic on board, Novo is building a multi-vendor AI infrastructure designed to tackle different aspects of the drug discovery challenge.

This multi-partner approach extends beyond AI models. In August 2026, Novo established a co-innovation hub with Amazon Web Services (AWS) in London, focused on applying agentic AI and cloud technologies to drug discovery. Agentic AI refers to AI systems that can autonomously plan and execute tasks with minimal human intervention, a capability that could accelerate the iterative process of testing and refining drug candidates.

• OpenAI Partnership: Novo uses Rosalind, a specialized GPT model, to analyze datasets and identify potential drug compounds from complex biological data.
• Anthropic Collaboration: Claude Science will support biological reasoning and help scientists understand human biology and drug mechanics through advanced language understanding.
• AWS Integration: A co-innovation hub applies agentic AI and cloud infrastructure to accelerate the entire drug discovery pipeline.
• Data Governance: All collaborations are designed with robust oversight and compliance standards to ensure responsible AI deployment in pharmaceutical research.

What's the Real Bottleneck in AI-Driven Drug Discovery?

While AI models can now design thousands of potential drug candidates in minutes, the actual validation of these candidates remains slow and expensive. A recent analysis of AI-driven protein drug development reveals that the true bottleneck has shifted downstream, from computational design to wet-lab testing.

Researchers can use generative AI and deep learning to create hundreds of protein binder candidates with atomic-level accuracy. However, each candidate must be physically synthesized and tested to verify that it actually folds correctly, binds to its target, and shows the desired biological activity. Traditional cell-based expression methods, which rely on living cells like E. coli or yeast, require weeks of work per candidate, including vector construction, cell culture, protein purification, and functional testing.

To address this constraint, scientists are increasingly turning to cell-free protein synthesis (CFPS), a technique that extracts the cellular machinery needed for protein production and uses it in a test tube. This approach can produce functional proteins in hours rather than weeks, dramatically accelerating the feedback loop between AI design and experimental validation.

A recent collaborative study demonstrated the power of this approach. Researchers combined Sino Biological's cell-free expression system with Cytiva's surface plasmon resonance (SPR) technology, a label-free method for measuring how tightly proteins bind to their targets. The team synthesized 200 different nanobody variants in just three hours, then screened all 200 candidates using SPR in an additional 4.5 hours, identifying 11 positive binders.

How to Accelerate the AI-to-Lab Feedback Loop

The integration of rapid computational design with efficient experimental validation requires a coordinated workflow. Here are the key steps that leading research organizations are adopting:

• Generative Design: Use AI models to create hundreds or thousands of protein candidates in silico, leveraging diffusion models and deep learning to navigate the vast protein sequence landscape with atomic-level precision.
• Cell-Free Synthesis: Deploy cell-free protein synthesis to bypass the weeks-long delays of traditional cell culture, producing functional proteins directly from DNA templates in hours.
• Label-Free Validation: Apply surface plasmon resonance or similar biophysical techniques to measure binding affinity directly from crude protein extracts, eliminating the need for time-consuming purification steps.
• Data Feedback: Feed the results from wet-lab experiments back into AI models to refine and retrain generative algorithms, creating a continuous cycle of improvement.

This compressed "build-test" cycle, which can now be completed in a single day instead of weeks, generates the high-quality experimental data needed to continuously improve AI models. The result is a virtuous cycle where AI designs better candidates, experiments validate them faster, and the models learn from real-world results.

What Does This Mean for the Future of Drug Development?

Novo's multi-vendor AI strategy reflects a maturing understanding within pharma that no single AI system will solve all drug discovery challenges. Different models excel at different tasks, from analyzing biological datasets to reasoning about molecular mechanisms to planning experimental workflows. By combining these capabilities with advances in rapid protein synthesis and biophysical validation, companies like Novo are positioning themselves to dramatically compress research timelines.

The collaboration also underscores the importance of responsible AI deployment in pharmaceutical research. Novo has designed its partnerships with Anthropic and AWS with robust data governance and human oversight, ensuring that AI tools are applied in line with ethical and compliance standards. This approach reflects the high stakes involved in drug discovery, where errors can have serious consequences for patient safety.

"AI's increasing capability brings with it the potential to compress a century's worth of biological and medical breakthroughs into a decade. By giving leading researchers access to safe, capable, and trusted frontier models, we can shorten research timelines, improve outcomes, and discover medicines and treatments that significantly improve human life," said Dario Amodei, co-founder and CEO of Anthropic.

Dario Amodei, Co-founder and CEO of Anthropic

For patients waiting for treatments for chronic diseases like diabetes and obesity, the implications are significant. If Novo and other pharma companies can successfully integrate AI across their research pipelines, the time from discovery to clinical use could shrink from 10 to 15 years to perhaps 5 to 7 years or less. That acceleration could mean millions of people gaining access to life-changing medicines years earlier than they otherwise would.