BenevolentAI

BenevolentAI builds biomedical knowledge graphs paired with machine learning to identify drug targets and predict how candidate molecules will behave in the body, aiming at diseases with few existing treatment options. Pharma and biotech researchers used the platform to mine scientific literature, clinical data, and genomic databases for connections a human team might miss, and the company's most visible deal was a target discovery collaboration with AstraZeneca on atopic dermatitis and chronic kidney disease.

That AstraZeneca partnership produced disappointing phase 2 results for its lead dermatitis candidate, and the fallout has defined the company since: repeated layoffs, a US office closure, leadership turnover pushed by founder Ken Mulvany, and a 2025 delisting from Euronext Amsterdam after BenevolentAI merged into Osaka Holdings and went private. The company has been shedding its earlier software-as-a-service ambitions and cutting costs to stretch its cash runway.

Under its current strategy, BenevolentAI is repositioning its technology into more modular, standalone tools that pharma partners can integrate into specific stages of drug development rather than selling a single end-to-end platform. It remains a cautionary example in the AI drug discovery field: sophisticated computational target identification still has to survive the same expensive, high-attrition clinical trial process as any other approach, and a strong knowledge graph doesn't guarantee a drug that works in patients.