The risks of using AI in pharmacy

Artificial Intelligence (AI) is rapidly transforming healthcare, including pharmacy practice. From clinical decision-support tools to prescription-checking software, patient-facing chatbots and workflow automation, AI offers the promise of improved efficiency and enhanced patient care.  

However, with these opportunities come new legal, regulatory, and professional risks, especially as pharmacists remain ultimately accountable for decisions influenced by AI. Understanding these risks, and how to manage them, is essential for any pharmacy professional engaging with AI-enabled systems. 

Accountability for AI-Assisted decisions

Even where AI is integrated into pharmacy practice, professional accountability remains unchanged. Guidance released by the General Pharmaceutical Council (GPhC) in April 2026 makes it clear that AI should support, not replace, professional judgement.  

Pharmacists are responsible for the care they provide and the decisions they make, whether or not AI tools are involved. 

Emerging case law from other professions provides valuable warnings. For example, in Ayinde v London Borough of Haringey [2025], the High Court found that legal professionals had a duty to verify caselaw authorities before relying on them.  

The court was highly critical of the submission of AI-generated material that had not been checked, emphasising that reliance on generative AI does not excuse professional failings. 

Such cases highlight several risks relevant to pharmacy:

• Over-reliance on AI outputs: professionals may defer to AI recommendations instead of exercising independent judgement
• Failure to identify inaccuracies or “hallucinations”: AI can generate plausible but incorrect information, which may go unnoticed without verification
• Automation bias: the tendency to trust automated suggestions over one’s own judgement
• Lack of independent verification: not checking AI-generated information against authoritative sources. 

While generative AI may not “hallucinate” a drug or prescription in the same way it can invent legal precedents, it seems likely that a court or the GPhC would take a similar approach in a pharmacy matter. Pharmacists are likely to be expected to verify AI outputs – and they will probably always retain ultimate responsibility for clinical decisions.

Regulatory and fitness to practise risks

In its April 2026 guidance, the GPhC reaffirmed that pharmacists must continue to exercise independent professional judgement, even when AI assists with medicines optimisation, prescribing, patient communications, or workflow automation. 

Some key regulatory risks for pharmacists include:

• Over-reliance on AI-generated outputs: decisions affecting patient care must never be delegated entirely to automated systems
• Lack of understanding of AI systems: pharmacists must be familiar with the capabilities and limitations of AI tools used in their practice
• Insufficient training: staff should ensure they receive appropriate training to identify potential inaccuracies, bias, or errors in AI-generated recommendations
• Transparency and governance: pharmacy providers must ensure they have clear policies, monitoring arrangements, and safeguards, especially where patient information is processed or clinical decisions are influenced by AI. 

Failure to address these risks may expose pharmacists and pharmacy providers to regulatory scrutiny. In particular, AI-related errors could lead to Fitness to Practise investigations around competence, professional judgement, or professional misconduct.  

Locum pharmacists may face additional challenges if they have limited time to familiarise themselves with different AI-supported systems at different workplaces. 

The GPhC’s expectation is clear: AI is a tool, not a replacement for professional expertise. Pharmacists must remain vigilant, ensuring that decisions remain rooted in clinical judgement.

Liability, indemnity and risk management

AI systems potentially complicate the investigation of liability for an incident. Where an AI system is alleged to have contributed to patient harm, disputes could arise as to whether the harm resulted from clinical negligence, technology failure or a combination of both.  

Injured patients could bring claims against pharmacists, pharmacy owners and AI providers, complicating liability and the apportionment of damages. 

The use of AI tools also affects indemnity arrangements. The GPhC requires pharmacists to have “adequate and appropriate” indemnity cover. The Insurance Act 2015 imposes a “duty of fair presentation”. This means that if the underwriter considers that material information (which may include the use of AI tools) was not disclosed when purchasing insurance, cover can later be withdrawn or reduced. 

To address these risks, pharmacists and pharmacy owners should:

• Consider AI when preparing to purchase or renew medical indemnity cover: take stock of current and planned AI-enabled systems and seek guidance from a broker or the intended insurer about how much detail you need to provide about AI use in your practice. Ensure that errors involving AI tools are within the scope of cover
• Maintain clear audit trails: document AI-assisted decisions and any steps taken to verify AI outputs
• Establish governance arrangements: implement clear policies and monitoring for AI-enabled services
• Provide comprehensive staff training: Equip staff to understand the limitations of AI and to identify potential risks. Empower them to exercise their own professional judgment
• Comply with data protection obligations: Ensure patient information is handled securely, in line with regulatory requirements. 

Regulatory bodies, including the GPhC, expect pharmacists to exercise independent professional judgement and to understand the limitations of AI systems. 

By proactively addressing liability, indemnity, and risk management considerations, and implementing robust practical safeguards, pharmacies can harness the benefits of AI while protecting patient safety and professional integrity.  

For now, the risk remains with the pharmacist rather than the technology. AI is a potentially powerful support tool, but it cannot replace the expertise, judgement, and responsibility of the pharmacy professional. 

If you have any questions or concerns about AI in pharmacy, contact Keystone Law