Everyone in pharma is asking what AI can do. Fewer are asking what it actually costs.
The token bill is becoming visible, but it is only the small part. The bigger cost is often hidden in data readiness, integration, compliance review, human oversight, training, governance and the organisational redesign needed to turn a clever pilot into a useful, safe, scalable capability.
For us, the key question is not “what is the price per token?” It is “what is the cost per useful, compliant, approved outcome?”
That shift matters. It forces us to connect AI investment to measurable value, tier use cases by risk, distinguish genuine productivity gains from hidden labour, and plan for what happens when adoption succeeds rather than when the pilot launches.
We’ve written a longer piece on the hidden cost of AI in pharma, and why financial discipline needs to sit alongside ambition if we want AI to scale responsibly.
Which question is hardest to answer in your organisation: value, risk, human work removed, cost that grows with usage, or what happens when adoption succeeds?
If you'd like to learn more and share your thoughts you can join our Digital Interest Group, please contact our Membership and Marketing Manager, mandip@pmsociety.org.uk