AI Is Expanding What’s Possible in Drug Discovery. Operational Friction Still Determines How Fast It Happens.

This blog post was written by Kevin Lustig, Cofounder and CEO of Scientist.com and Chris Petersen, Cofounder and CTO of Scientist.com.

For years, pharma treated AI as something to explore carefully, validate incrementally, and adopt selectively.

That posture is changing fast. Novo Nordisk has announced a broad partnership with OpenAI across drug discovery, manufacturing, and commercial operations. Anthropic has appointed Novartis CEO, Vas Narasimhan, to its board. The signal is clear: AI is no longer a side experiment. It is becoming part of how therapies are discovered, developed, and delivered.

At the scientific level, the benefits are already visible. AI helps teams analyze complex data, identify targets, generate hypotheses, and design studies faster than before. Across the industry, AI is being applied not only in discovery, but increasingly in clinical development, manufacturing, and post-approval activities, as well. The FDA has noted a significant rise in drug-development submissions that use AI components across the product lifecycle.

But execution has not accelerated at the same rate.

Most research teams still have to identify the right partner, confirm fit, define scope, and work through procurement, compliance, legal, and finance before work can begin. In too many organizations, ideas now move faster than the systems required to act on them.

That is the real bottleneck.

The biggest near-term gains from AI may not come from better prediction alone, but from improving the operational workflows that turn decisions into action.

When those workflows are connected, requests can be structured and routed faster. Supplier identification and comparison can happen in parallel. Legal, procurement, qualification, and finance steps do not need to restart from scratch with every new project. Study-start timelines shrink not because science changed, but because the process around the science did.

That shift matters across the full R&D lifecycle. As AI expands into development, clinical, manufacturing, and commercial functions, fragmented execution becomes even more costly. Faster decisions upstream only create more pressure downstream unless the workflow itself is integrated end to end.

When sourcing and execution are connected, another advantage appears: every request, decision, supplier interaction, and outcome becomes usable data. Over time, that data improves how future work is scoped, how partners are selected, and how quickly teams can move.

That is where durable advantage starts to compound.

If AI is raising the standard for how decisions are made in drug development, operations need to rise to the same standard.

The organizations that gain the most over the next decade will not be the ones with better models alone. They will be the ones that can consistently act on better decisions … across teams, across functions, and across the full pipeline.