Stepan Popov/ Alamy Stock Photo
Pharmacists risk becoming “liability sinks” in the period where regulation of artificial intelligence (AI) is in its infancy, Mark DasGupta, director of professional development at the Pharmacists’ Defence Association, has said.
Speaking at the Pharmacy Show, held in Birmingham on 11 October 2026, DasGupta described a “professional in the loop” concept of AI in healthcare, where “we see AI being used as a partner with a human professional and AI guiding and helping the professional to make their individual decisions”.
A “professional in the loop” means that the user understands how AI works, can challenge its outputs and can override if needed, he added.
However, DasGupta noted that human presence while AI is being used does not in itself guarantee that its use will be appropriate. AI cannot hold professional registration, cannot be held legally liable and cannot exercise professional judgement, he continued.
DasGupta said that, importantly, an AI “can’t apologise to the patient if they ever get something wrong”.
“The real risk for pharmacists here is that we become liability sinks because nobody else takes accountability for the use of AI and the outputs of AI,” he added.
“And that’s going to persist whilst regulation is catching up, because at the moment regulation is not ready for AI.”
The results of a Royal College of Pharmacy (RCPharm) survey, carried out in March 2026, revealed that just under two-thirds (63.8%, n=90) of respondents reported having used AI at work. The survey results also highlighted that almost three-quarters of respondents (74%, n=104) said that the use of AI in pharmacy practice gave them some cause for concern.
The PDA is one of several bodies that have produced guidance on AI for pharmacy teams. In January 2025, RCPharm published a policy statement on the use of AI in pharmacy.
In April 2026, the General Pharmaceutical Council (GPhC) published its position statement, which emphasised that pharmacists, pharmacy technicians and pharmacy owners must continue to meet the regulator’s standards at all times, including when using AI.
In September 2026, the GPhC was among other healthcare regulators who committed to jointly developing “high-level principles” around the regulation of AI use by health and social care professionals.
DasGupta welcomed the government’s acceptance of all 44 recommendations made by the National Commission into the Regulation of AI in Healthcare. He described this as “good news” and suggested that healthcare professionals should start seeing outputs in the next six months.
He said that he believed regulators are doing the best they can, but added: “Whether they have the capacity or even the capability to regulate AI, I’m not yet sure.”
He also said that, in his view, “one of the key issues is: we’ve got a small number of innovators in the field who are working to the attitude of ‘move fast and break stuff’”.
There is a need to balance public protection and patient safety without stifling commercial innovation, DasGupta said.
He continued: “I think in some areas that’s okay; I think you can move fast and break stuff. But I don’t want to see clinically important situations or politically important algorithms breaking people.”
Read more: ‘PJ view: Pharmacists must not be trapped by the promise of AI efficiencies‘
