Artificial Intelligence in Pharmacy: Evolution, Risk, and the Role We Need Next

The AI conversation in healthcare is picking up pace and rightly so. What was once considered experimental or speculative is now embedded in many of our clinical systems, often without fanfare. From machine learning algorithms supporting sepsis prediction to language models generating discharge summaries, we’re seeing real-world AI tools land in NHS workflows.

The pharmacy profession is entering a critical period of transformation. Not just because of AI, but because of the intersection of AI, data, automation, workforce pressure, and digital infrastructure. As with all periods of rapid change, there is both risk and opportunity. Somewhere in this tension lies the case for a new kind of role....

In pharmacy, though, the discussion has been slower to mature. There’s still a lot of misunderstanding, not just about what AI is, but what it isn’t, and what it might mean for the future of our profession. This blog explores where we really are with AI in pharmacy, where we’re heading, and why we urgently need to start formalising a role that reflects that reality: the AI Pharmacist.

Available now on Amazon

🤖 The Misunderstood Nature of AI

There is widespread confusion about what AI actually is, particularly in healthcare. The term is often used interchangeably with automation or analytics. In reality, AI refers to systems that mimic aspects of human intelligence, from pattern recognition to language generation.

Within AI, we find:

• Machine Learning (ML) – models that learn patterns from data and improve over time

• Deep Learning (DL) – complex neural networks used for tasks like image recognition and speech processing

• Natural Language Processing (NLP) – systems that extract meaning from human language

Healthcare examples include MedPaLM, trained on clinical questions and patient records, and IBM Watson, which was once heralded as the future of medical decision-making before its commercial retreat highlighted just how difficult clinical AI truly is.

In pharmacy, however, AI is already embedded in more subtle but important ways, from formulary logic and prescribing algorithms to demand forecasting and robotic dispensing units. We don’t always call it “AI”, but its footprint is growing.

🧠 What We Talk About When We Talk About AI

First, let’s be clear: AI doesn’t mean robots handing out methadone or ChatGPT prescribing chemotherapy. Artificial Intelligence in healthcare usually refers to a family of tools that can detect patterns, learn from large datasets, and generate predictions or classifications.

That includes:

• ML models trained on historical prescribing data

• NLP used to extract insight from free-text notes

• Computer vision models that scan radiology or pathology images

There is genuine promise here. According to the Royal Society, machine learning holds “the power and promise of computers that learn by example”, which can unlock real-world clinical value if embedded safely and meaningfully into care pathways (Royal Society, 2017).

But there is also hype... and hype leads to poor design, trust erosion, and unsafe implementation.

🏥 Where AI Is Quietly Reshaping Pharmacy Practice

Across the sector, AI is making its mark in distinct, operationally significant areas.

Examples include:

1. Clinical Decision Support (CDS) and EPMA

Advanced CDS modules are increasingly integrating probabilistic risk models, dose adjustment logic, and NLP-powered alerts. Combined with EPMA systems, they can identify anomalies, flag interactions, and support safer prescribing, but also risk alert fatigue and false reassurance if poorly designed.

2. Medicines Optimisation

AI tools can support identification of high-risk patients, automate elements of structured medication reviews, or identify patterns in non-adherence. In community pharmacy, this could evolve into personalised, proactive care, if data access barriers are resolved.

3. Operational Efficiency

Predictive stock control models, robotic dispensing, and automated scheduling algorithms are quietly reshaping how medicines move through supply chains. These are areas where human time is expensive but human judgement is less critical, making them prime targets for AI-supported optimisation.

4. Pharmaceutical Research and Industry

AI is accelerating drug discovery, repurposing, and adverse event detection. ML models can now screen compounds at a scale previously unimaginable, helping bring new molecules to trial faster and more efficiently.

As noted by Dash et al. (2019), the growing availability of structured clinical data, particularly from EHRs, has allowed predictive analytics and ML models to contribute to healthcare delivery at the system level.

👨‍⚕️ The Case for the AI Pharmacist

Despite this progress, pharmacists are rarely central to the design, evaluation, or governance of AI tools. That is a problem, because AI in medicines use isn’t neutral. It brings with it new risks: hidden biases, over-reliance on flawed models, or the slow creep of clinical deskilling.

We need pharmacists who can do more than just use these tools.

We need colleagues who can:

• Understand how an ML model works and when not to trust it

• Act as clinical safety advisors and escalation points in digital pathways

• Liaise with data scientists, software teams, and frontline clinicians

• Influence procurement decisions and digital strategy from a position of insight

Colleagues with a skillset of:

• Clinical insight to interpret outputs and spot flaws in automation

• Data literacy to engage with data-driven tools and their limitations

• Governance awareness to uphold standards, transparency, and safety

• Communication skills to liaise across clinical, technical, and strategic domains

In short, we need an AI Pharmacist, someone who brings the same professional lens to digital tools that we’ve always brought to medicines themselves: critical thinking, safety, and patient focus.

This isn’t a far-future vision. In practice, we’re already seeing this type of role appear under other names: Digital Pharmacy Lead, Clinical Informatician, Medicines Data Specialist. The next step is making this skillset explicit, structured, and teachable.

⚠️ Risk and Responsibility

It is important to be realistic: not all roles will survive untouched by AI. But history shows us that jobs don’t vanish, tasks within them evolve. Pharmacists who rely solely on technical dispensing, or resist digital up-skilling, may find their value diminished in a system driven by efficiency and automation.

At the same time, AI systems carry major risks:

• Bias embedded in training data

• Black-box decision making

• Scope creep from support to automation without accountability

It is tempting to treat AI like just another tool, but that downplays the systemic risks it introduces. Predictive models can replicate structural bias. Poorly implemented CDS tools can generate dangerous alert fatigue. Black-box systems can obscure accountability when something goes wrong.

As Brossard et al. (2022) point out in their review of data-driven hospital systems, AI’s value only emerges when paired with the right governance, workflow design, and human oversight. That’s exactly where pharmacists should be operating.

Pharmacists, especially those trained in governance, safety, and evidence, have a vital role to play as ethical gatekeepers of AI integration in medicines use and patient care.

📚 So What Can Pharmacists Do Now?

You don’t need to become a coder or data scientist, but you do need to build your AI literacy.

That includes:

• Understanding how different types of AI work (ML, NLP, LLMs)

• Getting familiar with tools already in your workplace

• Asking questions: how was this model trained? On what data? Who reviewed it?

• Exploring no-code platforms to build simple tools or workflows

• Contributing to AI projects, pilots, or procurement reviews within your organisation

Pharmacists are already well-placed, as a medicines experts, safety advocates, and systems thinkers. The AI Pharmacist role is less about knowing everything, and more about knowing enough to lead, question, and guide.

We need to move beyond hype and fear, and start building the profession we want to see.

That includes:

• Encouraging pharmacists to develop digital confidence and AI fluency

• Providing pathways for CPD, secondments, and academic collaboration in AI

• Involving pharmacy teams in the co-design and testing of AI tools

• Developing a shared vocabulary and governance framework across clinical and tech teams

• Advocating for regulatory clarity around AI-supported decision-making

Above all, we need a cultural shift: to see digital tools not as threats, but as extensions of our clinical practice.

📘 Want to Go Deeper?

If this topic resonates with you, whether you are a pharmacist, a digital health leader, or just someone interested in the intersection of tech and healthcare... you might want to read my latest book:

The AI Pharmacist: Redefining the Future of Healthcare and Pharmacy in the Age of Artificial Intelligence

It explores:

• What AI actually is (and isn’t)

• How it’s being used in pharmacy today

• What the future workforce might look like

• How to develop the digital skillset that’s becoming essential

Available now on Amazon 👉 The AI Pharmacist

(or search Kyle Cromey to see all books available in The Pharmacy Mastery Series)

Let’s not wait to be disrupted.

Let’s lead the change.