Patients booking cervical screening appointments through an AI chatbot respond positively to friendliness and choice-oriented language, but are put off by over-messaging, pushy reminders and blurred human-AI boundaries, according to new research from the University of Surrey.
In the study, published in Lingua, researchers found that how a health chatbot communicates matters as much as what it does. Using interviews with patients at a North London GP surgery and a survey of 300 people eligible for cervical screening via the NHS, the team analysed user perceptions of Asa, a generative AI receptionist that invites patients to book appointments via WhatsApp.
Patients who responded well to Asa described its tone as "friendly", "kind" and "not forceful" – and appreciated how the system integrated into their existing routines. Several noted that interacting with a named, female-presenting AI made it easier to disclose sensitive information, such as needing to cancel an appointment due to menstruation, compared with speaking to a male receptionist.
But the research also identified clear friction points. Many patients found follow-up messages sent within 24 hours intrusive, and described imperative phrasing such as "Let's book you in" as aggressive rather than helpful. For patients managing mental health challenges, neurodivergent conditions or demanding caring responsibilities, the pressure to respond quickly felt unfair.
Ethical concerns were the most consistently negative finding across the study. Patients raised worries about data security, impersonation and the blurring of human and AI boundaries. The chatbot's statement that users could "chat to me as if I am a real person" backfired for many, who read it as suspicious rather than reassuring.
Dr Doris Dippold, lead author of the study and Associate Professor in Intercultural Communication at the University of Surrey, said:
“Our analysis shows that anthropomorphism is not universally positive.
“Human-like features can build rapport - but when they clash with patients' expectations for transparency in a healthcare setting, they undermine exactly the trust the chatbot is trying to build."
The study suggests that design of healthcare chatbots should follow these key principles:
- Helping people achieve their goals
- Giving them control over decisions
- Responding appropriately to their needs
- Treating them with respect
- Ensuring fairness
- Being transparent about how the technology works.
The findings carry particular weight given that cervical screening uptake across the UK fell 5.3 per cent in 2023-24, with ethnic minority groups consistently underrepresented in screening programmes. The GP surgery at which Asa was trialled serves a highly diverse, socioeconomically deprived community in Islington, making equitable, accessible communication a priority.
The study highlights the importance of building rapport between patients and healthcare chatbots through communication that feels respectful, transparent and supportive.
Dr Dippold continued:
"Feeling seen, appreciated and emotionally supported is not a luxury feature in health AI - it is a condition of access.
“If patients disengage because a chatbot feels pushy or untrustworthy, the health service loses them entirely."
[ENDS]