Charles River to collaborate with Eli Lilly on TuneLab, adding nonclinical testing expertise to Lilly’s AI/ML drug-discovery platform

Eli Lilly said it is expanding its collaborative artificial intelligence and machine-learning drug discovery effort, Lilly TuneLab, by bringing in Charles River Laboratories. In a June 18 announcement, Charles River said it has joined the TuneLab platform, contributing nonclinical testing expertise intended to help improve how Lilly’s AI/ML work progresses toward real-world drug-candidate evaluation.

The companies positioned the collaboration around a practical bottleneck in drug discovery: turning promising model outputs into candidates that can be tested in preclinical settings. Nonclinical testing generally refers to experiments conducted before human trials, including studies used to gather evidence about potential safety and biological activity. By incorporating a contract research organization’s testing capabilities into its AI platform, Lilly appears to be aiming to tighten the feedback loop between modeling and experimentation.

Charles River Laboratories, which operates widely used drug-development testing services, framed the move as an effort to optimize Lilly TuneLab through “nonclinical testing expertise.” While the announcement did not lay out specific studies, timelines, or drug programs, the language suggests the collaboration is designed to strengthen the platform’s ability to evaluate how AI/ML-selected candidates perform when moved into nonclinical workflows.

The TuneLab concept is described by Lilly as a collaborative platform for AI/ML drug discovery. In general terms, these systems rely on large volumes of data and iterative training, but they can be slowed by gaps between computational findings and the experimental realities of preclinical testing. Lilly’s decision to partner with a provider like Charles River indicates it wants its AI outputs to be assessed with standardized testing capacity rather than relying solely on internal infrastructure or separate vendor processes.

For Charles River, the announcement is consistent with the broader market push toward integrated discovery and development services, where companies seek partners that can connect upstream discovery indicates with downstream translational evaluation. Charles River is a publicly traded company with a business built around running lab and study work for pharmaceutical and biotechnology clients, and joining an AI-focused platform indicates interest in helping shape how those clients operationalize data-driven discovery approaches.

Sector-wide, AI and machine learning are increasingly being used to identify therapeutic targets and prioritize molecule or formulation candidates. The challenge, companies have said across the industry, is translating computational predictions into robust experimental evidence. Partnerships that link discovery platforms with nonclinical testing services are one way firms try to reduce the time and iteration costs involved in moving from model hypotheses to preclinical data packages.

The announcement did not disclose what specific nonclinical testing modalities Charles River will support inside TuneLab, how results will be integrated into the platform’s modeling, or whether the collaboration is tied to particular therapeutic areas or molecules. It also did not provide financial terms, measurable performance targets, or expected impact on near-term revenue for either company. As a result, investors and industry observers will likely look for later details in company presentations, regulatory updates, or additional disclosures about participating programs and study scope.

Next, watch for any further communications clarifying how Charles River’s testing role will be structured, such as which types of studies are included and how often TuneLab will incorporate experimental results back into its AI/ML training cycles. It will also be important to see whether the collaboration is connected to specific Lilly research programs, and whether the companies provide evidence over time that earlier, nonclinical-informed optimization improves candidate selection outcomes.