Skip to main content
. 2025 Jun 12;28(1):e301787. doi: 10.1136/bmjment-2025-301787

Table 3. Attitudes towards LLM integration in mental health research among the total sample (n=714).

Item Strongly disagree Somewhat disagree Neither agree/disagree Somewhat agree Strongly agree
LLMs will become a standard tool in academic research within the next 5–10 years 14 (2.0%) 32 (4.5%) 56 (7.8%) 294 (41.2%) 318 (44.5%)
I would like more institutional support or training on responsible LLM use 33 (4.6%) 37 (5.2%) 96 (13.4%) 250 (35.0%) 298 (41.7%)
I have ethical concerns about the use of LLMs in research in general 15 (2.1%) 39 (5.5%) 101 (14.1%) 298 (41.7%) 261 (36.6%)
LLM use may compromise the scientific integrity or rigour of research outputs 17 (2.4%) 70 (9.8%) 125 (17.5%) 296 (41.5%) 206 (28.9%)
Researchers should be required to disclose LLM use in academic writing 15 (2.1%) 37 (5.2%) 92 (12.9%) 183 (25.6%) 387 (54.2%)
I have concerns that LLM use could affect how my academic work is evaluated (eg, by peer reviewers or funding bodies) 21 (2.9%) 76 (10.6%) 186 (26.1%) 264 (37.0%) 167 (23.4%)
I believe LLMs will reduce the need for certain types of research roles (eg, research assistants, copy editors, reviewers)? 74 (10.4%) 184 (25.8%) 149 (20.9%) 203 (28.4%) 104 (14.6%)

LLM, large language model.