AI drug discovery Market Outlook 2026–2036: Growth Drivers, Key Companies, Strategic Developments & Adoption Forecast — A Meticulous Next™ Foresight Brief

This Meticulous Next™ brief examines AI drug discovery, the use of generative models, phenomics and large datasets to identify drug targets and design molecules. It tests how fast AI-designed candidates move through mid and late stage trials and into approvals, and which participants capture value among AI developers, pharmaceutical partners, technology and compute providers and investors.

The pipeline is now large and the clinical record is short. An abstract for ASCO 2026, using data as of 1 December 2025, counted 117 AI-enabled assets from 63 companies in interventional trials, of which 60, or 51.3%, had completed phase 1 and 8, or 6.8%, had completed phase 2, and the median time from company founding to phase 1 entry was 6.5 years. Trackers put the number of clinical programmes at 173 to 175 in 2026, and none has received full FDA marketing approval. The conclusion is that AI has multiplied candidates entering the clinic, and approvals depend on phase 2 and phase 3 outcomes that have not yet arrived.

The most advanced proof point is rentosertib, a TNIK inhibitor for idiopathic pulmonary fibrosis from Insilico Medicine. In a phase 2a trial of 71 patients in China published in Nature Medicine in June 2025, the 60 mg group gained a mean 98.4 mL in forced vital capacity at 12 weeks against a loss of 20.3 mL on placebo, and the drug met its primary safety endpoint. A phase 3 programme was registered in July 2026, and the company reported dosing the first phase 3 patient in September 2026. The conclusion is that an AI-designed molecule has shown a signal against a hard target, and the small trial size means the phase 3 result will decide whether the signal holds.

Setbacks remain common. Nature Medicine reported in 2025 that AI-discovered drugs had experienced similar levels of phase 2 failure as other drugs, Recursion discontinued three programmes in May 2025, and Schrödinger discontinued SGR-2921 in August 2025 after two treatment-related deaths in a phase 1 trial. The conclusion is that AI shortens the path to the clinic more reliably than it raises the odds of success once a drug is in patients, and investors will price programmes on trial readouts rather than platform claims.

Capital and partnerships keep flowing. Isomorphic Labs raised a USD 2.1 billion Series B in May 2026, Generate:Biomedicines announced phase 3 studies of GB-0895 for severe asthma in December 2025, and large pharmaceutical companies including Lilly and Novartis have signed agreements with Isomorphic. The conclusion is that pharmaceutical companies are buying access to AI platforms and funding the best capitalised developers, which concentrates the market among a small group.

The brief treats AI drug discovery as a high-impact signal with a ten year horizon to 2036. It identifies which applications, regions and operators scale first and what each means for AI developers, pharmaceutical partners, technology providers and investors. It is a focused 106-page decision brief that presents an indicative trajectory rather than a segmented market model.

AI drug discovery uses machine learning to identify targets, design or screen molecules, predict their properties and prioritise candidates before laboratory and clinical testing.

Generative chemistry designs new small molecules, protein design tools create biologics such as antibodies, and phenomics platforms read cell images at scale to find biology that links targets to disease.

Foundation models trained on biological and chemical data support these steps, and cloud and GPU capacity sets the cost of training and running them.

Uses are oncology, fibrosis, immunology, respiratory disease and rare disease programmes, either in-house or in partnership with pharmaceutical companies.

The limits are phase 2 failure rates similar to other drugs, programme discontinuations, no approved AI discovered drug and data quality. The period turns on whether phase 2 and phase 3 results confirm higher success rates.

The AI drug discovery market, covering software platforms, discovery services, computing and partnered discovery programmes that use AI to identify targets and design drug candidates, is valued at USD 2.5–4.0 billion in 2026 and is projected to reach USD 25.21–40.34 billion by 2036, at a CAGR of 26%.

The 2026 base rests on software platforms, discovery services, computing and partnered programmes bought by pharmaceutical and biotechnology companies. Growth then follows clinical readouts, approvals and expansion of foundation models.

The forecast is sensitive to phase 2 and phase 3 results, financing and pharmaceutical partnering. Regionally, North America leads on developers and capital, Europe on platforms and partnerships, and Asia-Pacific on developers in China and Hong Kong and on pharmaceutical partners in Japan.

The base case assumes at least one AI-designed drug is approved by 2031 and phase 2 success rises modestly, giving the inflection in 2031. An accelerated case assumes several positive phase 3 results, bringing the inflection to 2029 and lifting the 2036 value to the top of the range. A delayed case assumes repeated late stage failures, holding growth at the bottom of the range.

Growth drivers

• First clinical proof points.
• Pharma partnerships and licensing.
• Large financing rounds.
• Foundation models for biology and chemistry.

Enablers

• Technology company investment.
• Pharma data and collaboration agreements.
• GPU capacity and cloud.
• Regulatory acceptance of AI in development.

Restraints and barriers

• Phase 2 failure rates similar to other drugs.
• Programme discontinuations and restructuring.
• No approved AI discovered drug.
• Data quality and model validation.

The direction of AI drug discovery is set by the need to raise clinical success and cut cycle time, regulation and evidence expectations, generative models and data scale, partnering among pharmaceutical companies, biotechnology companies and technology providers, and compute and data cost.

The key participants in AI drug discovery include Insilico Medicine Cayman TopCo, Recursion Pharmaceuticals, Inc., Alphabet Inc., Schrödinger, Inc., XtalPi Holdings Limited, Absci Corporation, Generate:Biomedicines, Inc., Iambic Therapeutics, Inc., Relay Therapeutics, Inc., NVIDIA Corporation, Eli Lilly and Company, Novartis AG, Sanofi, AstraZeneca PLC, Bristol-Myers Squibb Company, F. Hoffmann-La Roche Ltd, Verge Genomics, Inc., Xaira Therapeutics, Inc., Atomwise, Inc., Terray Therapeutics, Inc., Sumitomo Pharma Co., Ltd., Takeda Pharmaceutical Company Limited, Bayer AG, Johnson & Johnson, American Thoracic Society, American Society of Clinical Oncology, European Society for Medical Oncology, EMBL-EBI, Institute for Protein Design at the University of Washington, Stanford University, Massachusetts Institute of Technology, National Institutes of Health, U.S. Food and Drug Administration, European Medicines Agency, Medicines and Healthcare products Regulatory Agency, National Medical Products Administration, Pharmaceuticals and Medical Devices Agency, Biotechnology Innovation Organization and Pharmaceutical Research and Manufacturers of America.

The brief profiles representative participants in each archetype and assesses which are positioned to lead as AI drug discovery matures.

Insilico Medicine, Recursion, Schrödinger, XtalPi, Generate:Biomedicines, Iambic Therapeutics and Relay Therapeutics lead on clinical-stage assets, Alphabet, through Isomorphic Labs, and Xaira Therapeutics hold the best financed platforms, and NVIDIA supplies models and compute. Eli Lilly, Novartis, Sanofi, AstraZeneca and Bristol-Myers Squibb are the main pharmaceutical partners. Competitive intensity is high in 2026 and rises as readouts arrive.

Where value migrates.

In 2026 value sits with developers that have late stage assets and large pharmaceutical partners, because investors and partners pay for clinical results rather than platform claims. By 2031 it moves toward owners of approved AI-discovered drugs and platforms with proven success rates, because approvals reset expectations. By 2036 it settles in integrated design, testing and development platforms. Margins hold for developers that own assets and data and narrow for service providers that sell design at fixed fees.

The signals converge on a growing pipeline and a first late stage trial. A count of 117 assets in trials by December 2025, a phase 2a gain of 98.4 mL in lung function and a phase 3 start in September 2026 support growth, while phase 2 failure rates similar to other drugs, discontinuations and the absence of an approved drug limit the pace.

Structural limits show where progress could slow: phase 2 failure rates similar to other drugs, programme discontinuations and restructuring, no approved AI-discovered drug, and data quality. The brief therefore tracks four indicators: phase 2 and phase 3 readouts, first approvals, pharmaceutical partnership value, and financing of developers.

Oncology and fibrosis first, then immunology and respiratory disease, then wider use of generative biologics are the expected order of adoption.