Artificial intelligence is no longer a future issue for pharmacy; it is a current practice issue. Patient-facing tools, clinical decision support, documentation software, PMR automation and everyday search are already shaping decisions in pharmacies across the UK.
Yesterday, Sunday 11 October, Professor Mark DasGupta, Director of Professional Development at the PDA, presented “AI in Pharmacy: Who is really accountable?” in the Technology & Digital Health Theatre at the NEC Birmingham.
A question for the profession
The session examined the growing gap between what pharmacists are accountable for – patient safety, clinical decisions, professional standards – and what they actually control, such as the algorithm, the training data, system design and procurement decisions. The PDA’s concern is that pharmacists risk becoming the “liability sinks” while regulation catches up.
If AI influences a pharmacy decision and something goes wrong, who answers? The GPhC’s position is unambiguous: “Pharmacists and pharmacy technicians are personally accountable for their professional practice, even when AI is being used.”
The session also unpacked why a human being present is not the same as meaningful oversight and introduced the Professional-in-the-Loop (PITL) approach: an accountable professional who can understand a system’s purpose and limitations, challenge its outputs rather than merely approve them, override it where professional judgement requires, and justify the final decision independently. Mark set out some of the risks that should concern the profession, including hallucination, sycophancy, automation bias and deskilling through cognitive offload.
What pharmacists have told us so far
Interim results from the PDA’s AI snapshot survey show a profession that is open to AI but firmly safety-focused. Asked what safeguards should be required before AI tools are used in practice, members most often selected professional oversight (83%), training requirements (77%), human review of outputs (68%), regulatory approval (68%) and clear accountability arrangements (64%).
On the PDA’s own role, members were equally clear. 86% want the PDA to represent pharmacists’ concerns to employers, regulators and government, to provide guidance for members, and to support members where AI raises employment or professional concerns. A further 83% want the PDA to challenge unsafe implementation, and 81% to campaign for clear legal and regulatory safeguards.
A reminder: the PDA’s AI guidance
Members can already access the PDA’s professional guidance series on the safe and effective use of AI, covering how AI works and where it is used, the different types of AI and their risks, regulation, governance, data quality and bias, professional accountability and liability, practical steps to mitigate risk, when not to use AI, and future developments.
One core principle runs through all of it: AI can support decision-making, but it must not replace professional judgement.
The new guidance also sets out pragmatic questions to ask before using an AI system, including; “What is this system actually doing?” and “Could I justify the final decision independently?” If the answers are unclear, pause, verify and escalate.
Read the PDA AI guidance here
Learn more
Not yet a PDA member?
If you have not yet joined the PDA, we encourage you to join today and ask your colleagues to do the same.
Membership is FREE to pharmacy students, trainee pharmacists and for the first three months of being newly qualified.
Read about our key member benefits here.