Sanofi and OAO Use Multi-Agent AI to Crowdsource 1,300+ AI Use Cases from Employees and Scale Flagship Projects

Sanofi, working with the AI company OAO, uses a multi-agent system to drive AI adoption across the organisation rather than deploying isolated tools, according to the World Economic Forum's MINDS case study on the partnership, which was selected for the programme's second cohort in January 2026. OAO's agents operate across all levels of Sanofi's hierarchy, capturing insights from employees, translating their domain expertise into concrete use cases and helping leadership decide where to place focused AI investments. The challenge was that expertise in a large pharmaceutical company is spread across discovery, clinical, manufacturing and commercial teams that use different tools and decision frameworks, while traditional digital programmes tend to deliver isolated use cases that leave everyday work unchanged. The agents instead meet employees where they are, turning their input into initiatives and connecting local experiments with enterprise priorities, so that adoption becomes participatory rather than centrally imposed. The programme spans research and development, manufacturing and supply, and commercial operations. More than 1,400 employees have taken part, co-creating over 1,300 AI use cases; management has shaped 450 data-informed initiatives, and leadership has started more than 10 executive AI bets. Two flagships have been scaled enterprise-wide. In biologics discovery, a unified platform combining predictive and generative models with automated data ingestion and retraining has 11 models in production, and model build times have fallen from several months to under three. In commercial operations, an omnichannel system deployed across multiple markets and brands uses specialised models to recommend actions to teams engaging healthcare professionals, within their existing workflows. The case reports a 20% uplift in AI readiness and a single-digit commercial performance uplift with a 10:1 return on investment across deployed use cases.