TD Cowen: AI Cuts Drug Dev Timelines Up to 70%
Get the Health newsletter
Daily health & science — research, biotech, public health, the studies worth knowing. Free.
- TD Cowen's proprietary survey of 80 biopharma leaders found AI is compressing drug developers' preclinical costs and timelines by as much as 70%, per the firm.
- New drug development programs could grow by more than 10% over three to five years, fueling an estimated $1 billion in incremental spending on tech and labs.
- AI hasn't yet discovered a single drug that won FDA approval, and skeptics warn the industry drug failure rate could remain around 90% without better modeling of human biological diversity.
- Demand is surging for "in silico" simulation platforms that run thousands of virtual experiments in seconds, with software predicting drug interactions and pediatric dosing seeing the strongest upside by 2028.
- The Trump administration's push to reduce animal testing in biomedical research is expected to shift more work to computational tools, 3D human tissue models, and other non-animal alternatives.
- China's biotech buildup — cheaper labor and faster turnaround times — is already attracting billions in investment and threatening U.S. research efforts.
- Brendan Smith, TD Cowen's director of life sciences equity research, frames the strategy as "creating more shots on goal" through data generation that trains AI models to improve clinical success odds.
Why it matters: The 70% compression in preclinical timelines represents a structural shift in pharma R&D economics, redirecting roughly $1 billion in new spending toward AI platforms, simulation software, and 3D tissue models over three to five years. Wet-lab roles may stay relatively intact, but the 90% drug failure rate shows AI has yet to translate efficiency into actual FDA approvals — leaving investors and regulators exposed to whether the hype matches the clinical payoff.
