Major pharmaceutical companies are increasing investment and building internal AI capabilities to accelerate new drug discovery. The move centers on automated laboratories, closed-loop experimentation, and a strategic push to own core AI infrastructure.

Roche Leads with “AI Independence” Strategy

The “Lab in the Loop” Model

Roche and Genentech are implementing a “lab in the loop” approach where AI proposes hypotheses, robotics execute experiments, and results flow back to models for iteration. The objective, framed as “AI independence,” is to let algorithms generate and validate molecular ideas with minimal human bottleneck.

Boosting R&D Efficiency

Roche has committed roughly 2 billion CHF toward data and AI capabilities and set an objective of delivering 20 new molecular entities by 2030. Early implementations of tools such as Target Nexus are being used to prioritize targets and refine candidate selection, with pilot programs reporting improved decision quality and higher late-stage trial success in select programs.

Broader Industry Shift and AI Integration

Other large players are following suit. Novo Nordisk is adopting large AI models such as Claude for drug design tasks. Partnerships and service models are evolving: Twist Bioscience and GenScript are providing wet lab capacity to companies including Eli Lilly, which is building integrated AI platforms to orchestrate design and synthesis. The industry is moving from one-off collaborations to reworking R&D pipelines around automated, closed-loop dry-wet workflows.

What This Means for Drug Development

Widespread adoption of dry-wet closed loop systems promises faster candidate cycles and greater demand for high-throughput experimental verification. The next two to five years will be decisive as companies translate pilot gains into repeatable, commercial outcomes. Investors and researchers should watch how in-house AI stacks, lab automation, and partnership models reshape where value sits in the drug development ecosystem.