Pharmaceutical giant Sanofi and AI biotech company Owkin announced a multi-year partnership to co-develop next-generation AI agents that will autonomously perform complex drug research and development tasks. The collaboration, which builds on a previous €90 million strategic partnership dating back to 2021, represents a significant evolution in how artificial intelligence is being integrated into the drug discovery process.

What Are These AI Agents and How Do They Work?

The new AI agents, deployed through Owkin's platform called K Pro, are designed to function like intelligent assistants within Sanofi's existing workflows. Rather than simply analyzing data, these agents will autonomously handle multiple stages of pharmaceutical development. K Pro combines multimodal patient data, which includes diverse types of information ranging from genetic profiles to clinical outcomes, with specialized biological AI systems to support decision-making across the entire drug development lifecycle.

The platform aims to complement Sanofi's existing AI capabilities while providing what the companies describe as "competitive intelligence" to enable faster, more informed, and more precise decisions. This represents a shift from traditional AI tools that require human interpretation toward systems that can independently execute complex research tasks.

How Will These AI Agents Transform Drug Development?

• Early Discovery Phase: AI agents will assist in identifying promising drug targets and narrowing down which molecules warrant further investigation, reducing the time researchers spend on dead-end compounds.
• Clinical Development: The agents will help analyze patient data to identify which patient populations are most likely to benefit from a drug, enabling more targeted trial design and faster regulatory approval pathways.
• Competitive Analysis: K Pro will monitor the broader pharmaceutical landscape, providing insights into competitor activities and market positioning to inform strategic decisions about which programs to prioritize.

Thomas Clozel, CEO and co-founder of Owkin, emphasized the significance of embedding AI directly into pharmaceutical operations.

"Building on our collaboration with Sanofi, this marks a shift toward truly embedded AI," Clozel stated. "Owkin believes that, with K Pro, Sanofi can further harness agentic systems within their own workflows, unlocking the full value of their data to accelerate better decisions across drug development."

Thomas Clozel, CEO and Co-founder at Owkin

Why Does This Partnership Matter for the Pharmaceutical Industry?

The collaboration signals a broader industry trend toward what Owkin calls "Biological Artificial Superintelligence," a concept focused on using AI to solve complex biological problems that have historically stumped human researchers working alone. The five-year license agreement suggests that Sanofi views this technology as essential to its long-term competitive strategy, not merely as an experimental tool.

Emmanuel Frenehard, Chief Digital Officer at Sanofi, noted that the company is committed to implementing these purpose-built systems across its operations.

"Across Sanofi, we are continually investing in frontier AI solutions with the potential to accelerate and improve decision-making throughout the drug development lifecycle," Frenehard explained. "By implementing purpose-built agentic systems into our workflows, we aim to empower our teams to operate with greater speed, depth, and confidence as we continue to work to deliver transformative outcomes for patients."

Emmanuel Frenehard, Chief Digital Officer at Sanofi

The partnership builds on a track record of collaboration between the two organizations. Since 2021, Owkin and Sanofi have worked together on target identification in oncology and patient subgrouping, with the collaboration later expanding to include drug positioning for Sanofi's immunology pipeline. This new agreement represents the next phase of their relationship, moving from specific research applications to broader integration of AI agents across multiple therapeutic areas.

What Challenges Remain in AI-Driven Drug Discovery?

While AI agents show promise, the pharmaceutical industry still faces significant hurdles in deploying these systems reliably. One critical challenge is ensuring that AI systems provide accurate information rather than generating plausible-sounding but false outputs, a problem known as "hallucination" in AI research. Researchers at Binghamton University recently developed a novel approach to address this issue in biomedical AI applications.

The Binghamton team, led by Ahmed Abdeen Hamed and Luis M. Rocha, created a verification protocol that harnesses multiple large language models, or LLMs (AI systems trained on vast amounts of text to understand and generate human language), to cross-check medical information. Rather than relying on a single AI system, the protocol uses seven different LLMs and requires them to reference authoritative medical databases before providing answers. The LLMs then "vote" on the correct answer.

The results were striking: across more than 10,000 experiments, 76.85% of answers were supported by at least four of the seven LLMs, and the remaining 23.15% were supported by at least two. Critically, no unmatched or hallucinated terms appeared in the results, suggesting that this verification approach could significantly improve the reliability of AI systems used in drug discovery and clinical decision-making.

"The new workflow is incredible because it can verify anything from a biomedical point of view, biological knowledge with disease and genetics, translational knowledge from diseases to treatments and clinical trials, and also from a healthcare point of view with symptoms and treatments," Hamed noted.

Ahmed Abdeen Hamed, Research Fellow at Binghamton University

The Binghamton protocol demonstrates that accuracy in AI-driven drug discovery may depend less on building a single, more powerful AI system and more on creating verification mechanisms that cross-reference multiple AI approaches. This insight could inform how companies like Sanofi implement and validate the AI agents being developed through their partnership with Owkin.

As pharmaceutical companies increasingly embed AI agents into their research workflows, the ability to verify and validate AI-generated insights will become as important as the speed and scale these systems provide. The Sanofi-Owkin collaboration represents the industry's confidence that agentic AI is ready for real-world deployment, while research like Binghamton's work highlights the ongoing need for robust safeguards to ensure these powerful tools deliver trustworthy results.