Patients Frustrated by AI Receptionist at GP Surgery
Patients at an Oxfordshire GP surgery have raised concerns about an AI receptionist, highlighting challenges with healthcare access, communication and digital technology.
Introduction
Artificial intelligence is becoming an increasingly common feature of healthcare, with GP surgeries exploring new technology to manage calls, streamline administrative tasks and improve access to medical services. However, the introduction of an AI receptionist at a GP surgery in Oxfordshire has sparked frustration among some patients, raising questions about whether automated systems are making healthcare easier or creating additional barriers.
Newbury Street Practice in Wantage introduced its AI-powered Andi telephone system in July 2026. Since its launch, some patients have reported difficulties communicating with the virtual receptionist, describing their experiences as frustrating and, in some cases, a barrier to accessing care.
While the practice says most feedback has been positive, the concerns highlight an important issue for the NHS: how can healthcare providers benefit from artificial intelligence without leaving patients feeling unheard or excluded?
What Is the AI Receptionist at Newbury Street Practice?
Newbury Street Practice, based at Wantage Health Centre in Oxfordshire, introduced the Andi AI system to handle incoming telephone calls. Under the arrangement, all calls to the surgery pass through the automated system, which can then direct patients to a human receptionist when needed.
According to the practice, the technology is intended to improve access, increase administrative capacity and give staff more time to support patients who cannot or do not wish to use digital services.
The practice's website describes Andi as a system designed to listen to callers, understand their needs and help them obtain assistance more quickly.
AI receptionists are increasingly being explored across healthcare settings because GP practices face substantial administrative demands. Answering telephone calls, directing enquiries and helping patients navigate appointment processes can consume considerable staff time.
Automation may help manage these tasks. However, its success depends on whether the technology can understand different callers, recognise their needs accurately and provide a straightforward route to human assistance.
Patients Describe Frustration With Automated Calls
Despite the practice's stated aims, some patients have expressed dissatisfaction with their experiences of the new system.
Patient Jo Rogers described the AI receptionist as a barrier to obtaining care. She reported waiting in a telephone queue for approximately 20 minutes before the system determined that it could not help her.
Her experience raises concerns about the additional effort patients may face when an automated system cannot resolve their enquiries.
For someone trying to contact a GP surgery about a medical concern, a lengthy wait followed by an unsuccessful interaction can be discouraging. Patients may already be worried about their symptoms or uncertain about the appropriate next step.
Another patient, 65-year-old Tricia Lockhart, reported that the system mistakenly interpreted her call as being about pregnancy. She expressed frustration at the apparent misunderstanding, highlighting the importance of accurate communication in healthcare administration.
These accounts illustrate a potential weakness of automated telephone services. Although AI can process speech and identify common requests, misunderstandings may occur when callers use unexpected language, explain complex circumstances or encounter problems with speech recognition.
For patients, the key issue is not simply whether the technology works in general. It is whether it works reliably when they need help.
Why AI Receptionists Can Be Challenging for Patients
Introducing artificial intelligence into GP reception services can offer practical benefits, but it also presents challenges that practices need to consider.
1. Communication Difficulties
Not every patient communicates in the same way. Hearing difficulties, speech differences, accents, background noise and unfamiliarity with automated telephone systems can affect how accurately an AI tool understands a caller.
A system that misunderstands a patient's request may create unnecessary delays or direct the person towards an inappropriate service.
2. Limited Access to Human Support
Some patients prefer speaking directly to a receptionist, particularly when discussing sensitive matters or explaining a situation that does not fit a standard category.
If patients must interact with AI before reaching a member of staff, they may feel that an additional obstacle has been placed between them and the care they need.
3. Digital Exclusion
Digital healthcare can be particularly challenging for people who have limited confidence with technology, disabilities or difficulties using automated services.
The Newbury Street Practice has acknowledged that some patients, particularly those less confident with technology, have found the transition more difficult.
Healthcare providers must ensure that modernisation does not unintentionally disadvantage the very people who need accessible services.
4. Confidence and Trust
Patients need confidence that their concerns will be understood and directed appropriately. Repeated misunderstandings or lengthy interactions may weaken that trust.
Even when an AI system is intended to improve efficiency, patients may judge its effectiveness by how easily they can get assistance.
The GP Surgery's Response to Patient Concerns
Newbury Street Practice has defended the introduction of Andi, stating that the majority of feedback received during the system's first month was positive.
According to a practice spokesperson, 79% of feedback was positive during that initial period.
The practice said the technology was introduced to improve access, increase capacity and allow staff to spend more time supporting patients who cannot or choose not to use digital options.
It also acknowledged that some patients had struggled with the change, explaining that new systems require time to settle. The practice said it regularly reviews its processes and patient feedback to determine what further support may be needed.
The spokesperson also stated that the technology was being used by more than 1,300 practices across the UK and had undergone extensive NHS assurance checks.
These assurances suggest that the practice sees AI as a way to support its existing services rather than replace the human element of patient care. Nevertheless, patient experiences demonstrate why ongoing evaluation is essential.
Positive overall feedback does not necessarily mean that every patient can use the system successfully. Practices must also understand the experiences of those who encounter difficulties.
Local Healthcare Pressures Add to the Debate
The introduction of the AI receptionist has also attracted attention from local MP Olly Glover, who represents the Didcot and Wantage constituency.
After hearing from residents, Glover contacted the surgery and acknowledged that GP practices are attempting to streamline their administrative processes.
He also highlighted wider pressures on primary care services, particularly in areas experiencing significant housing growth.
Growing communities can place additional demands on GP surgeries, increasing the number of patients seeking appointments and administrative support. When demand rises faster than available resources, practices may explore technology to manage workloads.
However, automation alone cannot resolve every underlying problem. Its effectiveness depends on staffing, service design, patient needs and the availability of appropriate healthcare resources.
The challenge for GP practices is to balance operational efficiency with accessible, reliable patient support.
Can AI Improve Access to NHS GP Services?
Artificial intelligence has the potential to improve healthcare administration when it is implemented carefully.
An effective AI receptionist could help manage routine enquiries, direct calls to the appropriate department and reduce pressure on reception staff. This could allow human employees to spend more time helping patients with complex requests.
However, these benefits depend on several important safeguards.
First, patients should have a clear and accessible way to speak to a human receptionist when the AI system cannot understand or resolve their enquiry.
Second, practices should monitor call waiting times, unsuccessful interactions, patient complaints and feedback from people who find automated services difficult to use.
Third, systems should be designed with accessibility in mind, including the needs of older people, patients with hearing or speech difficulties and those who are less confident using technology.
Finally, patients should receive clear information about how the system works and what to do if it fails to address their concerns.
AI should support healthcare professionals and administrative staff, not become an unnecessary barrier between patients and medical services.
Conclusion: Putting Patients at the Centre of Healthcare Technology
The concerns raised about the AI receptionist at Newbury Street Practice demonstrate the challenges of introducing artificial intelligence into everyday healthcare services.
Although the surgery reports that most initial feedback was positive, experiences shared by patients such as Jo Rogers and Tricia Lockhart show that automated systems can create frustration when communication breaks down.
As more GP practices explore AI-powered reception services, the priority must remain clear: patients need reliable, understandable and accessible ways to contact their healthcare providers.
Technology can help improve efficiency, but its value should ultimately be measured by whether it makes healthcare easier to access.
For the NHS, the future of AI receptionists will depend not only on technical performance but also on listening to patients, addressing concerns and ensuring that human support remains available when it matters most.
Sources
- BBC News, Local Democracy Reporting Service, published 14 October 2026.
- Local Democracy Reporting Service (LDRS), reporting on local government, public services and community issues.
- Newbury Street Practice, practice information and communications concerning its AI telephone receptionist, as referenced in the supplied news report.