AI-Designed Drug Candidates Reach Clinical Trials at a Pace That Would Have Been Unthinkable Five Years Ago

AI-native drug discovery companies — Isomorphic Labs (DeepMind's drug discovery spinout), Recursion Pharmaceuticals, and a growing roster of similarly AI-first biotech startups — now have multiple AI-originated drug candidates in active clinical trials, a milestone that represents genuine acceleration of the traditionally multi-year process of moving from target identification to a compound ready for human trials. The core technical advance enabling this compression is protein structure prediction and interaction modeling building on the AlphaFold lineage, which lets researchers computationally screen far more candidate molecules against a target protein before committing to the expensive, slow process of physical lab synthesis and testing. The clinical trial success rate for AI-originated candidates remains the genuinely unresolved question the industry is watching closely: computational discovery can dramatically speed up and cheapen the early screening stage, but it doesn't change the fundamental biology-driven failure rate of drug candidates in human trials, where most candidates across the whole pharmaceutical industry — AI-discovered or not — still fail at some stage between early trials and approval, making it too early to say definitively whether AI-originated candidates will show a meaningfully different overall success rate than traditionally discovered ones. The business model taking shape across AI drug discovery companies splits into two main approaches: platform companies like Recursion that aim to discover and develop their own drug pipeline directly, versus companies positioning themselves as a discovery engine that partners with and licenses candidates to traditional pharmaceutical companies for the clinical development and commercialization stages, a split that reflects genuine disagreement in the sector about whether AI-native companies should try to become full pharmaceutical companies themselves or specialize in the discovery stage alone. Traditional pharmaceutical companies have responded by building internal AI discovery capability and forming partnerships with the AI-native firms rather than treating them purely as competitors, with several major pharma companies now running both in-house AI discovery teams and external partnerships in parallel, reflecting an industry-wide view that AI-assisted discovery is becoming a standard part of the drug pipeline rather than a novelty separate track.