Key takeaways
- AI is most likely to take on routine, repetitive contact: practice information, simple requests and sorting messages, not the judgement and care that patients most value.
- Adoption in general practice is growing but uneven: 31% of GPs in England use AI tools, against 20% in Scotland and 9% in Northern Ireland[1].
- Patients are shifting channels: 30.8% last contacted their practice online in 2026, up from 16.8% in 2024[2], but many still need the phone and the desk.
- Earlier waves of digital change did not shrink the general practice workforce, which grew 27% from 2019 to 2023[3]; roles changed instead.
- The skills that grow in value are care navigation, handling complex and distressed patients, safeguarding awareness and running the systems themselves.
The short answer
AI is unlikely to replace medical receptionists as a group in the foreseeable future, but it is likely to change what many of them do each day. The parts of the job that are routine and repetitive, such as answering the same questions about opening hours, explaining how to order a repeat prescription or sorting incoming messages, are the parts AI tools handle best. The parts that rely on judgement, empathy and local knowledge, such as calming a frightened parent, spotting a safeguarding concern or working out what a confused older patient actually needs, are the parts they handle worst.
That makes the honest answer a practical one: some tasks will move, some roles will change shape, and practices that plan the change carefully will keep their best people and use them better.
Which tasks are likely to move
| Task | Likely role for AI | Likely role for people |
|---|---|---|
| Practice information (hours, location, services) | Answer most questions at any time | Update the information and handle exceptions |
| Explaining processes (fit notes, repeat requests, results) | Explain the steps and route the request | Handle anything unusual or disputed |
| Sorting online requests and messages | Suggest categories and urgency for staff to check | Make the routing decision under protocol |
| Urgent symptoms | Recognise keywords and signpost urgent help | Follow up, judge, escalate to clinicians |
| Booking appointments | Possible where systems allow and rules are simple | Complex bookings, multiple needs, vulnerable patients |
| Distressed or angry patients | Recognise and hand over quickly | Listen, de-escalate, resolve |
| Safeguarding concerns | Limited | Notice, record and report |
| Desk and building | None | Greeting, specimens, wayfinding, emergencies |
Notice that most rows are shared rather than handed over completely. Even where a tool does most of the work, someone has to set it up, check what it does, handle what it passes on and fix it when it gets something wrong.
What the adoption evidence shows
In a 2025 Nuffield Trust and RCGP survey of 2,108 GPs, 28% said they currently use AI tools in clinical practice[1]. Use was higher in England than elsewhere in the UK: 31% of GPs in England, compared with 20% in Scotland and 9% in Northern Ireland[1]. Among GPs using AI, 57% used it for clinical documentation and note-taking and 44% for administrative tasks[1].
Source: Nuffield Trust, 2025[1]
Show the data
| Item | Value |
|---|---|
| England | 31% |
| Scotland | 20% |
| Northern Ireland | 9% |
So far, then, AI in general practice is mostly helping clinicians with notes and admin rather than replacing front-desk staff. Reception-facing tools, such as automated phone assistants and message sorting, are newer and less widely used, and practices adopting them are still learning where they work well.
Patients are changing how they make contact
Patients are moving online, and many are comfortable with AI. In 2026, 30.8% of patients said they last tried to contact their GP practice online, up from 22.4% in 2025 and 16.8% in 2024[2]. In the government's public attitudes tracker, 40% of UK adults said they use AI chatbots for personal purposes at least once a month[4].
The flip side is that most patients still use the phone or the desk, and some always will: people without reliable internet, people with communication needs, people who are very unwell and those who simply prefer a voice. Any change that makes it harder for them to reach a person would make access less fair. That is one of the strongest reasons reception roles will remain.
Where AI could make reception better
The most promising uses of AI at reception are not about removing people but about removing the worst parts of the job. The 8am queue, where every receptionist spends an hour repeating the same answers while urgent callers wait, is the clearest example. A tool that answers routine information questions at the busiest times and outside opening hours can shorten that queue and leave staff with the calls that actually need them.
An AI Employee such as Ava, for example, is designed to cover peaks and quiet hours for practice-information questions and to pass patient matters to the reception team rather than resolve them alone. Used that way, the tool becomes another member of the front-desk rota with a narrow, well-defined job, and the people around it do the work that needs judgement.
What earlier changes suggest
General practice has been through several waves of digital change: online booking, electronic prescriptions, online consultation forms and the NHS App. Each was expected by some to reduce admin workload. In practice, demand kept rising and the work changed shape. NHS England reported that the general practice workforce grew by 27% from 2019 to 2023 to meet rising demand and an ageing population[3]. Online requests created new admin tasks even as they reduced some calls.
There is no guarantee the next wave will follow the same pattern, but it is a useful reminder that technology in general practice has usually redistributed work rather than removed it.
Skills that grow in value
- Care navigation: knowing every role and service well enough to send each patient to the right place first time.
- Complex and distressed callers: listening, calming and solving problems that do not fit a script.
- Safeguarding awareness: noticing what a system cannot see, such as who is with the patient and how they behave.
- Running the systems: keeping practice information accurate, reviewing what the tools say and fixing errors.
- Data and reporting: turning call, request and workload data into decisions about rotas and processes.
- Coordination: following up on handoffs from automated tools so nothing falls between systems and people.
These skills are harder to recruit and take longer to learn than answering routine questions. Pay should reflect them. Typical receptionist pay of £21,000 to £25,000 a year[5] does not always recognise the expertise of experienced staff who do this work well.
The risks of getting it wrong
There are real risks if practices treat AI as a straightforward replacement for staff. The most serious is patient safety: a tool that misses an urgent symptom, or a handoff that nobody picks up, can cause harm. The second is fairness: if the only easy route in is automated, patients who struggle with technology lose out. The third is trust: patients currently rate their contact with practices more positively than before, with 72.6% describing their last contact experience as good in 2026[6], and a poorly configured tool could reverse that.
- Cutting staff before the tool has proved itself, leaving nobody to handle what it passes on.
- Letting a tool give answers nobody at the practice has reviewed.
- Hiding the route to a person, which frustrates patients and increases complaints.
- Ignoring governance, including data protection and clinical safety checks.
- Failing to tell staff what is changing, which drives experienced people to leave.
What practices should do now
Measure the work
Count what reception actually does for two weeks so you know which tasks are routine and which need judgement.
Decide what you want freed time for
Before choosing any tool, agree where released hours will go: care navigation, the online request list, document work or more time for complex patients.
Involve staff early
Ask receptionists to help choose and test tools. They know which questions patients ask and where things go wrong.
Invest in the human skills
Train staff in care navigation, conflict handling and safeguarding, and recognise those skills in pay and job titles.
Keep a human route open
Make sure every patient can reach a person by phone or at the desk, whatever tools you add.
Example (fictional): A practice with nine receptionists adds an automated assistant for routine information calls. Instead of reducing posts, it trains two receptionists as care navigators, moves one into document workflow and keeps a person on the phones at every opening hour. A year later the team is the same size, overtime has fallen and the receptionists spend more of their day on patients who need help to find the right service.
If you work on reception
If you are a receptionist wondering what this means for you, the most useful step is to build the skills that tools find hardest. Ask to be trained in care navigation, take on the role of checking what an automated system tells patients, learn how to produce simple reports from the phone system, and look at routes into supervision, coordination or secretarial work. Staff who understand both patients and systems will be the ones practices most need.
It also helps to be the person who speaks up when something is wrong. If an automated answer confuses patients, or a handoff keeps arriving without the details you need, say so and suggest a fix. Practices value staff who make new systems work, and that experience is exactly what employers will look for as these tools spread.
Frequently asked questions
Will AI replace medical receptionists?
It is more likely to change the job than remove it. Routine information and sorting tasks are the most likely to move to AI tools, while judgement, empathy, safeguarding and handling complex patients remain human work.
Which reception tasks are best suited to AI?
Answering repeated questions about practice information and processes, routing simple requests and helping sort messages, always with a clear route to a person and staff checking what the tool does.
How can receptionists future-proof their careers?
Build skills in care navigation, handling difficult conversations, safeguarding, data and reporting, and supervising automated systems, and look at routes into coordination, secretarial and supervisory roles.
References
Nuffield Trust (2025). How are GPs using AI? Insights from the front line (opens in a new tab) ↑
Ipsos (GP Patient Survey provider for NHS England) (2026). 2026 GP Patient Survey results released (opens in a new tab) ↑
NHS England and DHSC (2023). Delivery plan for recovering access to primary care (opens in a new tab) ↑
GOV.UK (Department for Science, Innovation and Technology) (2024). Public attitudes to data and AI: Tracker survey (Wave 4) report (opens in a new tab) ↑
National Careers Service (GOV.UK) (2026). Receptionist job profile (opens in a new tab) ↑
Ipsos (GP Patient Survey provider for NHS England) (2026). 2026 GP Patient Survey results released (opens in a new tab) ↑