Will AI replace pharmacists? What the 2026 data shows

No, AI is not going to replace pharmacists, at least not on any timeline the official data covers. The U.S. Bureau of Labor Statistics (BLS) projects pharmacist jobs to grow 5% from 2025 to 2035, faster than the average for all jobs, and the reasons it gives for where those jobs will and won’t be have little to do with AI.

But pharmacists are not safe from change either. Their work is unusually information-heavy, and Microsoft’s data puts them among the most AI-applicable jobs in all of healthcare, slightly ahead of software developers. This article covers the official numbers, what AI can and can’t do in a pharmacy task by task, the bigger threat that has nothing to do with chatbots, the arguments on both sides, and what to do if you are a pharmacist, a technician or a student deciding whether pharmacy school is still worth it.

The official numbers: high exposure, steady growth

About 325,200 people work as pharmacists in the U.S., and the BLS expects about 342,300 by 2035, a gain of 17,100 jobs. It projects about 12,500 openings a year, most of them to replace pharmacists who retire or change careers. The median pharmacist earned $140,910 in May 2025, close to three times the $50,980 median for all workers.

In August 2026 the BLS also rated 831 occupations for AI exposure, splitting them into four roughly equal groups from low to very high. Pharmacists and pharmacy technicians are both in the high group. Here is how they compare with nearby health jobs:

Two things stand out. First, exposure and growth point in different directions: nurse practitioners and medical transcriptionists share the same rating, yet one is among the fastest-growing jobs in the country and the other is shrinking. Second, pharmacists sit in the middle of the pack. Their growth ranks 227th of 831 occupations, and 61st of the 206 in the high-exposure group.

Why pharmacist jobs are moving, not disappearing

The most useful part of the BLS outlook is the explanation behind the 5%. It describes two different job markets:

• Hospitals and clinics are hiring. Demand is projected to rise as pharmacists move beyond dispensing and join care teams to manage patients’ medications.
• Retail is flat. Demand is expected to be limited in drugstores, supermarkets and mass merchandisers as the industry consolidates and more people fill prescriptions online or by mail.

Pay follows the same split. In May 2025, hospital pharmacists earned a median $157,290, while those in pharmacies and drug retailers earned $132,940.

The retail squeeze is already visible. A study published in Health Affairs in December 2024 by a team led by Jenny Guadamuz and colleagues at USC’s Schaeffer Center tracked 88,930 retail pharmacies that operated between 2010 and 2020 and found that 29.4% had closed by 2021. Independent pharmacies were more likely to close than chains, and pharmacies in mostly Black and Latinx neighborhoods were at greater risk than those in White neighborhoods. More pharmacies closed than opened from 2018 to 2021.

None of that is AI. It is reimbursement pressure, consolidation and mail order. But it matters for this question, because the parts of pharmacy that are shrinking are the same high-volume, dispensing-focused jobs where automation fits best.

What AI can and can’t do in a pharmacy

The best task-level evidence comes from Microsoft, which analyzed 200,000 anonymized conversations with its Copilot assistant and scored 785 U.S. occupations by how much of their work AI was used for, and how well. Pharmacists scored 0.28, ranking 119th of 785. That is higher than any of the other health jobs in our table, and higher than pharmacy technicians (260th), family doctors (334th) and registered nurses (460th). For comparison, software developers rank 120th, as we covered in will AI replace software engineers?

The reason is simple: much of what pharmacists do is explaining information, and that is what chatbots are built for. In Microsoft’s data, requests to explain medical information to patients and advise people on medical issues were completed successfully about 97% of the time, and AI covered a large share of each activity. Requests to present research or technical information were also among the most common.

The activities AI barely touched tell the other half of the story. Treating patients, preparing mixtures (compounding), conferring with other clinicians about a patient and replenishing stock were rare in the data or scored very low for scope. And the lowest completion rate among a pharmacist’s activities came on verifying personal information, about 72% when AI was doing the work. Those are the physical, accountable and checking parts of the job.

In practice, a pharmacist’s week splits like this:

• Most exposed: information and paperwork. Answering routine drug questions, writing patient leaflets, summarizing guidelines, drafting prior authorizations and handling insurance messages. AI drafts these quickly.
• Partly exposed: checking and screening. Flagging interactions, allergies and dose problems. Software has done some of this for decades, and AI makes it better, but a pharmacist still has to decide what an alert means for this patient.
• Least exposed: judgment, hands and accountability. Verifying the final prescription, counseling a confused patient, calling a prescriber about a risky dose, giving vaccines, compounding and managing complex medication plans in a hospital.

Anthropic, which measures what its Claude models are actually used for at work, gives a pharmacy-adjacent example that shows why theory and practice differ: authorizing drug refills and passing prescription information to pharmacies is a task AI could in principle speed up, yet it had not seen Claude used for it. Legal rules, software requirements and human checks keep such tasks in human hands. We saw the same pattern for doctors.

Can AI replace pharmacists? Where it still fails

The strongest argument against handing pharmacy to AI is how it fails. In a study presented at ASHP’s Midyear Clinical Meeting in December 2023, pharmacists at Long Island University tested the free version of ChatGPT on 39 real questions sent to their drug information service.

• Only 10 of 39 answers were judged satisfactory.
• The rest included answers that didn’t address the question, were inaccurate or were incomplete.
• Asked for references, ChatGPT gave them in 8 answers, and every one contained references that did not exist.
• Asked whether the COVID-19 drug Paxlovid interacts with the blood pressure drug verapamil, it said no interaction had been reported. The two can interact, and taking them together may drop blood pressure too far.

Current models are better than the 2023 free version, and a small single-site study is not the final word. But the lesson has held up: AI is fluent and often right, and when it is wrong it sounds just as confident. In dispensing, one confident mistake can harm a patient, which is one reason every state requires pharmacists to be licensed: a person, not a tool, answers for what leaves the counter.

Which pharmacy jobs are most and least at risk?

Most exposed:

• High-volume retail verification roles, where the day is mostly checking scripts at speed. This is the work most likely to be split between automation, central fill and remote verification, and it sits in the shrinking part of the market.
• Drug information and medical writing roles, whose output is mostly text. AI drafts much of it well, though a pharmacist still checks it.
• Routine technician tasks such as data entry, insurance calls, counting and inventory. Pharmacy aides, the most routine role, are projected to shrink slightly.

Least exposed:

• Clinical pharmacists in hospitals and clinics, who manage medications as part of a care team. This is where the BLS sees growth.
• Ambulatory care and specialty pharmacists, for conditions like cancer, HIV or anticoagulation, where decisions are complex and patients need ongoing contact.
• Pharmacists who give vaccines and run clinical services, which require hands, licenses and face-to-face trust.
• Pharmacy leaders and informatics pharmacists, who decide how AI tools are bought, tested and used safely.

Our own index puts health jobs at the safe end of the scale. A typical registered nurse’s week scores 16 out of 100 for AI exposure in our test’s example jobs, compared with 70 for a software developer and 96 for a data entry clerk:

What about pharmacy technicians?

Technicians are in a better position than many people assume. The BLS projects pharmacy technician jobs to grow 6% to 2035, from 471,200 to 501,300, with about 44,700 openings a year. Part of the reason is the aging population and rising chronic disease. The other part is directly related to the pharmacist’s changing role: as pharmacists do more patient care, technicians are expected to take over tasks pharmacists used to handle, such as collecting patient information, transferring prescriptions and checking other technicians’ work.

The median technician earned $45,750 in May 2025. Technicians who move into those advanced tasks, or into hospital, sterile compounding or informatics roles, will be less exposed than those whose job is mostly data entry and phone calls.

Will AI replace pharmacists? Both sides

The question comes up often on Reddit and pharmacy student forums. Here are the strongest arguments each way, useful whether you are deciding on a career or preparing for a debate.

Why some people think AI will shrink pharmacy jobs

• Pharmacists rank 119th of 785 jobs for AI applicability, above almost every other health job.
• Drug information and patient education, a large part of the job, is exactly what chatbots do well.
• Robots and mail-order pharmacies already handle counting and packaging in many places.
• Retail chains are closing stores and cutting costs, and AI gives them another way to do more with fewer staff.

Why the evidence says it won’t replace them

• The BLS projects 5% growth and 12,500 openings a year to 2035.
• Every state requires pharmacists to be licensed, and someone has to own the error when a tool gets it wrong.
• AI still makes confident, dangerous mistakes on drug questions, including missed interactions and invented references.
• Pharmacists’ work is shifting toward patient care, vaccines and medication management, which AI barely touches.

On balance, AI is more likely to change what pharmacists spend their day on than to remove the job. The bigger risk to a pharmacy career right now is the retail business model.

Will AI replace pharmacists in the future?

Ten years out, expect a pharmacy where AI and automation do most of the first-pass work: screening orders, drafting answers to drug questions, handling insurance paperwork and spotting problems in refill patterns. Pharmacists will spend less time counting and typing and more time verifying, counseling, adjusting doses and working with prescribers.

The pressure on workload may even ease. In the 2024 National Pharmacist Workforce Study, which surveyed more than 5,100 pharmacists, 73% rated their workload high or excessively high, up from 66% in 2014. Tools that take paperwork off the counter could help, as long as employers use the time for patient care rather than cutting staff.

Is pharmacy school still worth it?

For students, the honest answer is that it depends on the path. The same workforce study found pharmacist unemployment at just 2.8% in 2024, down from about 5% in 2019, and about 90% of pharmacists said their work is meaningful. But 53% of pharmacists work in community settings, the part of the market the BLS expects to stay flat.

Anthropic’s March 2026 labor market study adds a broader caution: it found tentative evidence that hiring of 22- to 25-year-olds into AI-exposed occupations has slowed slightly. That makes the first job matter more. Students who aim for residencies, clinical specialties, ambulatory care or informatics will be in a stronger position than those planning a career behind a retail counter.

If you are weighing other health careers, these from our index combine strong demand with lower AI exposure:

How to AI-proof a pharmacy career

The rule is the same in every job we cover: let AI do the production and keep ownership of the judgment. Here is how each type of task our test measures is affected:

For pharmacists specifically, five moves matter most:

1. Move toward patient care. Medication management, chronic disease clinics, anticoagulation and vaccines are where the growth is.
2. Get clinical depth. A residency, board certification or specialty is the clearest hedge against both AI and retail consolidation.
3. Use AI tools, and know where they fail. Being the pharmacist who catches the chatbot’s missed interaction is more valuable than the one who ignores the tools.
4. Learn informatics. Someone has to choose, test and govern the AI in pharmacy systems, and that person should understand both drugs and data.
5. Practice the human parts. Counseling worried patients and persuading prescribers to change a dose are among the least automatable skills in healthcare.

For a broader plan, read our guide to the skills AI can’t replace, or see the 25 most AI-proof jobs if you are still choosing a direction.