Summary: As traditional health services struggle with long waiting lists and rising costs, a large international survey reveals a marked shift in public trust toward artificial intelligence. The study of nearly 31,000 adults across 35 countries found that 41% of adults in the UK — and 61% globally — would be comfortable using ChatGPT-style tools for mental health support.
AI’s perceived benefits — a non-judgmental tone, round-the-clock availability and a private-feeling conversation — make these systems attractive to many. At the same time, researchers warn that AI cannot replace trained professionals and raise concerns about long-term effects on memory, learning and other cognitive functions if these tools substitute traditional education and health care.
Key facts
- “Counselor” shift: 41% of UK respondents would use AI for counseling, likely influenced by long wait times for conventional mental health services.
- Companionship: Three-quarters of people globally, and over half in the UK, said they would use AI as a friend or companion, drawn by adaptive tone and perceived privacy.
- Trust in medicine: 45% of respondents worldwide — and 25% in the UK — said they would trust AI to act as their doctor; trust was higher in countries with expensive or hard-to-access health care.
- Education concerns: One quarter of UK adults would delegate teaching their children to AI. Researchers caution this could encourage “prompt-focused” learning at the expense of deeper knowledge retention.
- Biological risks: Neuroscientists warned that excessive replacement of traditional learning with AI-supported search could reduce engagement of memory systems such as the hippocampus, potentially harming long-term memory and spatial learning.
Source: Bournemouth University
More than four in ten adults in the UK say they would use ChatGPT for mental health support, new research suggests.
The study, led by Bournemouth University, surveyed almost 31,000 adults in 35 countries about their willingness to delegate socially important roles to AI large language models such as ChatGPT. In addition to the findings above, researchers reported:
- One quarter of UK adults would be comfortable delegating the role of teaching their children to AI.
- Globally, 45% of respondents said they would trust AI models to act as their doctor.
- Three quarters of survey participants said they would use an AI chat tool as a companion or friend.
The research appears in the journal AI and Society. Dr Ala Yankouskaya, Senior Lecturer in Psychology at Bournemouth University and lead author, said the team set out to understand how people evaluate generative AI tools when asked to carry out important roles in daily life.
AI for mental health support
Forty-one percent of UK participants and 61% of respondents globally said they would be willing to use AI for counselling services. The authors suggest that long NHS waiting times and barriers to accessing care may push people toward AI as a faster, private source of support.
Dr Yankouskaya noted that while AI can provide an immediate listening ear, the responses she reviewed often avoided firm diagnoses and were sometimes vague. That cautious language reflects developers’ intent not to provide clinical diagnoses, and it underlines that AI is not a substitute for qualified mental health professionals.
Familiarity with existing NHS chatbots and other automated services may also be normalising AI use for health-related conversations.
AI as a teacher
A quarter of people in the UK, and roughly half of respondents worldwide, said they would trust AI to take on teaching roles for children. The research team highlighted this as particularly worrying. They caution that relying on AI for education might encourage a style of learning focused on crafting prompts rather than building durable understanding.
Researchers also raised concerns about potential long-term effects on brain development if traditional learning practices are replaced by frequent use of search and AI assistants — for example, reduced stimulation of the hippocampus, the brain region associated with memory and spatial awareness.
AI as a doctor
Across the full sample, 45% said they would trust AI to perform the role of a doctor; in the UK this figure was 25%. Higher acceptance appeared in countries where health care is expensive or inaccessible, which suggests people in those contexts may rely on technology for quick answers. The authors caution, however, that AI’s conversational design could prioritize keeping users engaged and relaxed rather than prompting them to seek urgent services when needed.
AI as a companion
The role that respondents were most willing to delegate to AI was companionship. Over three quarters of people globally and more than half of respondents in the UK said they would use ChatGPT-like systems as companions. The researchers attribute this to the adaptive, empathetic-feeling tone of generative language models and their non-judgmental approach, which many users interpret as supportive and private.
Dr Yankouskaya and colleagues concluded that as AI becomes more integrated into daily life, public awareness of how these systems operate and what they cannot do must increase. In particular, caution is needed when considering AI for education and other roles that shape cognitive development.
Key questions answered
A: Accessibility and fear of judgment are major factors. People facing long waits for a therapist or worried about stigma may turn to AI for immediate, private conversational support. AI can feel non-judgmental and consistent, which some users find comforting.
A: Not fully. AI can offer general information and emotional support, but its language is often intentionally cautious and it lacks the nuance, clinical judgment and safety safeguards of trained human professionals. In crises or complex cases, human intervention remains essential.
A: It could be problematic. Researchers warn about “cognitive outsourcing”: if people stop exercising memory and deep learning because AI finds answers for them, this may weaken the brain systems involved in retention and flexible thinking over time.
Editorial notes
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full by the editorial team.
- Additional context was added by staff to clarify implications.
About this AI and psychology research news
Author: Steve Bates
Source: Bournemouth University
Contact: Steve Bates – Bournemouth University
Image: The image is credited to Neuroscience News
Original research: Open access. “Who lets AI take over? Cross-national variation in willingness to delegate socially important roles to artificial intelligence” by Ala Yankouskaya, Mohamed Basel Almourad, Magnus Liebherr, Fahad Beyahi, Guandong Xu & Raian Ali. AI & Society
DOI: 10.1007/s00146-026-02858-5
Abstract
Who lets AI take over? Cross-national variation in willingness to delegate socially important roles to artificial intelligence
Delegating socially significant roles to artificial intelligence is an emerging reality, but public attitudes vary across countries and contexts. This study applied a structural model to a large cross-national dataset (30,994 individuals in 35 countries) to examine how cognitive appraisals, affective dispositions and contextual factors jointly shape willingness to delegate roles such as companion, mental-health advisor, teacher and doctor to AI.
Results reveal a consistent hierarchy of delegation preferences: companionship was most commonly entrusted to AI, followed by mental health advisor, teacher and doctor. Cognitive appraisals, especially trust in online information, were the strongest predictors of willingness to delegate. Affective factors, like anxiety and loneliness, had more specific effects. Gender differences were present, with women generally less willing than men to delegate across roles. National context also mattered: differences between countries persisted even after accounting for individual predictors.
Overall, the findings indicate that delegating socially important roles to AI reflects a blend of personal attitudes and broader societal conditions, suggesting the need for careful public education and policy consideration as AI systems become more prevalent in everyday life.