Kenneth Mulvany stared at the pharmaceutical industry’s brutal statistics. A decade from target to medicine. Billions per approved drug. Nine out of ten candidates are failing.

Meanwhile, biological data exploded exponentially. Genomics. Proteomics. Clinical trials. Scientific literature doubles every few years. The industry was drowning in data but starving for knowledge.

Human researchers couldn’t see patterns hidden in billions of data points. Couldn’t simultaneously consider millions of variables. Couldn’t interrogate entire biological systems at once.

Mulvany, who’d sold his previous biotech, Proximagen, for $553 million, believed artificial intelligence could change everything—not replace human scientists but augment them with computational power, turning years into hours.

In 2013, when AI drug discovery remained nascent, he founded BenevolentAI.

This represents why BenevolentAI, a London-based AI-driven drug discovery company, has won the 2026 Global Recognition Award for developing the Benevolent Platform™ that identified baricitinib as a COVID-19 treatment in 90 minutes of compute time (subsequently clinically validated and approved), advanced multiple pipeline drugs, including BEN-8744 achieving positive Phase Ia safety data, and established partnerships with AstraZeneca, Merck KGaA, and Eli Lilly. Founded in 2013, BenevolentAI achieved a €1.5 billion valuation in Europe’s largest SPAC acquisition (April 2022) and developed a proprietary knowledge graph rebuilt every few weeks that integrates billions of biological relationships.

Technical Innovation and Architecture

Traditional drug discovery operates on a hypothesis-driven, single-target approach. Scientists propose a mechanism. Test it. Fail. Try another. Sequential. Slow. Constrained by human cognition.

BenevolentAI’s Benevolent Platform™ completely changes the paradigm.

The Benevolent Knowledge Graph represents billions of relationships between diseases, drugs, proteins, genes, biological processes, clinical trials, and scientific publications. Not isolated facts. Relationships. Context. The way complex biological systems actually work.

The graph rebuilds from scratch every few weeks, ingesting the latest data from around the world. AWS cloud infrastructure processes petabytes, identifying novel connections invisible to humans.

The AI suite includes Target identification models that analyze genomic, transcriptomic, and proteomic data to identify dysregulated pathways. Drug repurposing algorithms scan approved molecules to identify new indications based on their mechanisms of action. Molecular design tools predicting drug-like molecules with desired properties. Iterative learning systems in which scientists’ decisions feed back, creating virtuous cycles.

Chief Scientific Officer Dr. Anne Phelan: “Our suite of algorithms enables us to predict targets based on foundational data and knowledge graph…This allows the intentions of scientists to be recorded within the platform, but also creates a virtuous cycle of learning and decision making.”

The COVID-19 repurposing of baricitinib exemplifies platform power.

Pandemic emerges. Traditional discovery would require years of identifying repurposing candidates through sequential hypothesis testing.

BenevolentAI’s AI: 90 minutes of compute time. Plus 3 days of human expert validation. Identified baricitinib—a JAK inhibitor approved for rheumatoid arthritis—could reduce both viral entry and dangerous inflammatory response.

Clinical trials validated the prediction. Regulatory approval followed, and millions of patients benefited years sooner than with traditional approaches.

The in-house pipeline demonstrates sustained productivity: BEN-8744, an orally administered PDE10 inhibitor for ulcerative colitis. AI identified PDE10 inhibition as a novel approach, but recognized that CNS side effects have plagued previous PDE10 inhibitors. a novel approach, but recognized that CNS side effects have been. Scientists deliberately designed a peripherally restricted molecule to avoid those side effects. March 2024 Phase Ia reported positive safety data validating AI-guided design.

BEN-2293 for atopic dermatitis progressed through Phase I/II targeting out-licensing—preclinical candidates for ALS and glioblastoma under development.

In-house wet-lab facilities integrate computational predictions with experimental validation—synthesizing predicted molecules, testing binding, assessing biological activity. Not pure software. Integrated discovery platform.

Market Strategy and Leadership

Kenneth Mulvany brings a pedigree as a serial entrepreneur. Founded Proximagen, developing CNS drugs. Sold to Upsher-Smith for $553 million in 2012. 20+ years building biotech companies.

His 2013 founding of BenevolentAI predated the AI drug discovery hype by years. Early-mover advantage, establishing platform maturity and dataset accumulation before competitors emerged, 2018-2020.

September 2016 ~$100 million Series B funded AI supercomputer and hired Dr. Jackie Hunter—former GlaxoSmithKline SVP—as BenevolentBio CEO. Big Pharma’s credibility. Industry connections.

April 2022: €1.5 billion valuation Euronext Amsterdam SPAC listing. Europe’s largest SPAC acquisition. €225 million gross proceeds.

Pharmaceutical partnerships validated platform: AstraZeneca collaboration (April/May 2019) on chronic kidney disease and idiopathic pulmonary fibrosis. Merck KGaA partnership. Eli Lilly’s collaboration with equity investment.

These generate upfront fees, milestone payments, and potential royalties on commercialized drugs discovered using BenevolentAI technology.

December 2024 boardroom changes: Mulvany returned as Executive Chairman. “Return to original mission,” refocusing on core platform-driven discovery.

February 2025 announced delisting (completed March 2025). Transition to a private structure, eliminating “substantial costs associated with maintaining public status” and diverting resources from core activities and restructuring, extending runway into 2027, through organizational and budgetary changes.

Strategic evolution prioritizing long-term platform development over short-term public market pressures.

Industry Impact and Future Vision

The pharmaceutical crisis: Despite the exponential growth in biological data, the industry cannot translate this information into proportional increases in the number of new medicines.

10-15 years. $2.6 billion per drug. 90% failure rates. Patients with rare diseases and complex conditions are waiting years for treatments that might never arrive.

Human researchers cannot see patterns hidden in billions of data points. Cannot simultaneously consider millions of variables. Traditional hypothesis-driven approaches cannot systematically interrogate entire biological systems.

BenevolentAI enables hypothesis-free discovery by having AI interrogate entire biological systems simultaneously and generating thousands of testable predictions, which humans prioritize based on confidence scores.

Drug discovery accelerates from years to months.

The COVID-19 baricitinib repurposing: 90 minutes of computation + 3 days of validation. Subsequently approved. Millions benefited years faster.

The platform identifies novel targets for complex diseases, including ulcerative colitis (BEN-8744), atopic dermatitis (BEN-2293), ALS, and glioblastoma, with breadth across therapeutic areas.

Pharmaceutical partnerships with AstraZeneca, Merck KGaA, and Eli Lilly validate the commercial value—the discovery collaboration model scales across an unlimited number of partners without incurring linear costs.

For patients: Accelerated medicines for underserved conditions. Rapid response to emerging health crises.

For pharma: Transformed R&D productivity. Hypothesis-driven sequential investigation becomes AI-enabled parallel exploration of thousands of targets simultaneously.

For healthcare: Systemic shift from reactive treatment development to proactive target identification for underserved populations.

This recognition is for fundamentally reimagining how humanity translates biological knowledge into life-saving treatments. For processing exponentially growing data that humans cannot comprehend, for bringing hope to millions suffering from diseases that traditional approaches cannot adequately address.

The data deluge became actionable knowledge. The decade became months. The 90% failure rate faces computational prediction. Welcome to BenevolentAI—winner of the 2026 Global Recognition Award.