It probably won't surprise anyone that AI has come up in almost every conversation we've had while researching ManuPharma 2027.

What has surprised us slightly is how practical those conversations have become.

There's much less interest in talking about AI as a big, abstract idea. Manufacturing leaders want to know what they can actually do with it.

Can it help predict equipment failures?

Can it make root-cause analysis faster?

Can it help people make sense of all the manufacturing data they're already collecting?

Could it even help an operator who has a straightforward question about an SOP?

One person we spoke to described that last use case as almost like 'first aid' for operators - giving someone an immediate answer rather than having them search through documents or track down a colleague.

Another manufacturing leader talked about wanting to get to 'real AI'. Their point was that collecting manufacturing data isn't really the problem anymore. The opportunity is using AI to do something useful with it - predictive maintenance and root-cause analysis were two examples.

That distinction has stuck with us.

For a long time, the industry conversation has focused on what's possible with AI.

Manufacturers seem ready for a different conversation: What is actually working?

Where has somebody implemented AI in a GMP manufacturing environment? How difficult was it to validate? What changed as a result? And was the improvement big enough to justify the effort?

There's clearly enormous potential for AI in manufacturing. But the people we've been speaking to don't seem particularly interested in potential on its own. They want evidence.

These are exactly the kinds of practical AI conversations we'll be continuing at ManuPharma 2027, with a focus on what's actually being implemented in manufacturing - and what others can learn from it.

Register for ManuPharma 2027 here.