Aipedia: Pharmacists and pharmacies
The pharmacist who signs off a clinical check is personally accountable for it, even when AI screening did the flagging. As screening is automated, the risk moves from missing a warning to trusting a silent system.
Part of: AI in pharmacists and pharmacies
In short
Patient medication record (PMR) and dispensing systems may use AI to flag interactions, doses, duplications and contraindications, and some automate steps of the clinical check. The GPhC position statement of April 2026 names clinical checks as an AI use and says pharmacists remain personally accountable and must check outputs. Standards 4 and 5 require knowledge of the tool and independent professional judgement. A pharmacists' defence association asks for extra care where automation reduces the visibility of decision-making, giving PMR automation as an example. A missed interaction claim will usually be brought against the pharmacist and the owner, not the software.
Traditional PMR alerts follow fixed rules. Newer systems add AI to rank alerts by likely significance, suppress low-value warnings, compare doses against a patient's history, and in some cases mark a prescription as cleared so that only flagged items reach the pharmacist. In hub-and-spoke and high-volume settings, this can mean the pharmacist sees a short exception list rather than every item. The time saved is real, but the pharmacist may no longer see the reasoning behind what was cleared.
The relevant duties are:
August 2026 guidance from a pharmacists' defence association, reported by the Pharmaceutical Journal, says AI should not be used where it reduces decision clarity or undermines accountability, and singles out automation in PMR systems as a change to clinical processes that needs extra care.
Illustrative scenario: a community pharmacy upgrades its PMR to a system that auto-clears items it rates as low risk. A new prescription for a commonly used antibiotic is cleared for a patient on a long-term medicine with which it interacts, because the long-term medicine was supplied by a different pharmacy and is not on the record. The pharmacist reviews only the exception list and does not see the item. The patient is admitted with a predictable adverse reaction.
Other failures include alert suppression that hides a serious warning among dismissed ones, dose checks based on out-of-date weight or kidney function, and staff treating a silent system as a safe one. Doctors face the same issue with prescribing alerts; see AI in prescribing and medication alerts.
A claim for a missed interaction is usually framed as a negligent clinical check, so it lands on the pharmacist who completed it and on the owner as employer or contractor. Pointing to the software rarely helps the pharmacist directly, though the owner may have contractual recourse against a supplier depending on the contract. Article 32 of the Pharmacy Order requires each registrant to hold appropriate indemnity; employed pharmacists may rely on employer cover for the duties of their job, but the GPhC says not to assume it is adequate. Whether a policy responds to an AI-assisted check depends on the wording, and a pharmacy policy may not mention AI at all.
A serious miss may also be reported to the GPhC, and a fitness-to-practise investigation is a separate matter from the negligence claim. State schemes do not pay for regulatory defence, and employer arrangements may not either, so check whether your own arrangement includes it; see regulatory investigation cover.
These steps reduce the risk:
See also dispensing robots and accuracy checks, chatbots for medicines queries, does pharmacist indemnity cover AI and the pharmacy AI hub.
Apex arranges indemnity for pharmacists and pharmacy owners. A broker reviews how your cover responds to a claim arising from an automated or AI-assisted clinical check.
Not in the GPhC's eyes unless you have applied your own judgement. The position statement says pharmacists remain personally accountable and must be able to justify their decisions, so relying on an auto-clear without understanding it is hard to defend.
Possibly, but that depends on the contract between the supplier and the pharmacy owner and is not settled law. In practice the patient usually claims against the pharmacist and the owner, who may then look at recourse against the supplier.
You need to understand its intended use, limitations and biases, as the GPhC position statement puts it, and be trained to use it. You do not need to know the code, but you should know what it checks, what it ignores and when it clears items without you.
Yes. Standard 8 requires pharmacy professionals to speak up when things go wrong, and the GPhC statement links AI use to the duty of candour. Owners need these reports to monitor tools as the statement requires.
Tell us about your firm and the work you do. A broker will review how your current PI treats AI-assisted work, or arrange new cover. No fee for a quote. Or call 0117 325 0027.
Start a PI quote Request a callbackApex Insurance Brokers Limited is authorised and regulated by the Financial Conduct Authority. Registered in England and Wales, company number 07014570. This page is general information about how AI is used in the profession and how insurance may respond. It is not legal, regulatory or insurance advice on your circumstances, and it does not guarantee that cover will be available or on what terms.