AI GP receptionist ‘can’t understand Yorkshire accent’
Patients in Rotherham have struggled to be understood by a new artificial intelligence receptionist, EMMA, used by some GP surgeries to triage calls, according to health watchdog Healthwatch Rotherham. The system, deployed nationwide to cut phone queues, reportedly fails to respond well to regional accents and speech impediments, causing frustration for callers with broad Yorkshire accents and prompting some vulnerable or elderly patients to visit surgeries in person instead.
Healthwatch Rotherham manager Kym Gleeson said accents vary significantly across South Yorkshire and the AI can struggle to understand enquiries, though she acknowledged the system speeds up call handling overall. QuantumLoopAI, which makes EMMA, said the tool supports 17 languages plus English, transfers calls to human staff when it cannot cope, and has received "overwhelmingly positive" feedback, with over 90% of patients reporting improved service. Healthwatch is passing feedback to Healthwatch England and the Department of Health and Social Care, while NHS South Yorkshire Integrated Care Boards said it is working with Healthwatch on guidance for practices using such AI telephone systems.
- AI GP receptionist EMMA struggles with Yorkshire accents, Healthwatch says
- Some elderly, vulnerable patients now visit surgeries in person instead
- Maker QuantumLoopAI cites 90%+ positive feedback; humans available on request
New here? Start with this
GP surgeries are part of the NHS and provide first-contact care for many health problems. Some use automated telephone systems to answer calls and help direct patients to the right service; EMMA is one such system, made by the company QuantumLoopAI.
Healthwatch Rotherham is a local watchdog that gathers patients’ experiences and raises concerns with health services. Its manager, Kym Gleeson, says some callers have found that EMMA struggles with regional accents or speech impediments, while the company says the system can pass calls to human staff and has received positive feedback.
The issue matters because people need to be able to contact their surgery to seek care, including those who are older or vulnerable. NHS South Yorkshire’s health bodies are working with Healthwatch on guidance for practices using these telephone systems.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
The system's defenders emphasise that reducing phone queue backlogs significantly benefits the majority of patients facing genuine delays in accessing GP services, and note that EMMA includes safeguards such as transferring calls to human staff when comprehension fails. They argue that a 90% satisfaction rate demonstrates overall effectiveness, that technological solutions improve through deployment and refinement rather than in isolation, and that abandoning the efficiency gains achieved for most patients would be disproportionate when the system already includes fallback mechanisms for problematic interactions.
The case against
Critics argue that healthcare systems must serve all patients reliably regardless of accent or speech difference, and that a system which fails users with common regional accents in its area of deployment—forcing vulnerable patients to make unnecessary in-person visits instead—has fundamentally failed basic accessibility standards. They contend that accent and speech variation are not marginal edge cases but ordinary human diversity deserving equal consideration, and that satisfaction metrics may mask genuine barriers experienced by the least able patients, making robust accent recognition an equity issue essential to fair healthcare access rather than a technical refinement.