Johnson & Johnson started with nearly 900 GenAI projects. But in time, it became clear: only 10–15% of them were driving 80% of the value.
The lesson? It’s not about doing more AI. It’s about doing the right AI — with clear ownership and real outcomes.
1. From Central Oversight to Strategic Ownership
- The company initially relied on a centralized AI governance model.
- Now, it has shifted responsibility to business units like R&D, Supply Chain, and Commercial — where decisions are closest to impact.
- This isn’t a return to silos — it’s a mature handover of ownership.
2. High-Impact Use Cases
- Drug discovery optimization
- Supply chain risk prediction and planning
- Rep Copilot tool: AI that supports field reps in real-time doctor interactions
3. Governance Reimagined
- AI is now governed at the business level — with support, not interference.
- The focus has shifted from “exploring ideas” to “scaling what proves itself.”
- It’s not about how many use cases you launch — it’s about which ones move the needle.