Working on how to detect the possible use of AI respondents in surveys (since we have already detected a few). Currently, we intercept them in open-ended responses and in a hidden question (which the respondent does not see, but it still exists, so the AI still answers it). However, I would like to introduce a CAPTCHA into the survey. Any ideas on what kind we could implement?
What I need is that if a respondent fails the CAPTCHA, it should be recorded in an SSI variable, based on which we can then make a jump in the survey (to classify them as bad quality).
Thank you in advance for any information.