What do AI chatbots do after users decline therapeutic redirection for suicide thoughts?

arXiv:2610.21253v1 Announce Type: cross Abstract: Safety evaluations of artificial intelligence (AI) chatbots focus on therapeutic redirection, statements urging users who disclose suicide thoughts to seek help from crisis lines or other human sources. How chatbots respond when users refuse that help is unclear. This cross-sectional audit ran 2,550 scripted multi-turn conversations in September 2026. Each conversation opened with a disclosure of suicide thoughts at 3 levels of severity, answered every referral with up to 10 standardized refusals and then changed the subject. All 2,550 conversations included a referral, 2,542 (99.7%) in the first reply. However, mean persistence ranged from 0.01 to 9.94 refusals across chatbots. Newer versions of all 5 chatbots tested in 2 versions persisted less than older versions (each P < .05). Compared with passive ideation, a disclosure of intent was associated with 1.26 more refusals sustained (95% CI, 1.06-1.46) and higher odds of renewed referral after the topic change (odds ratio, 10.07; 95% CI, 7.43-13.66). Chatbots that persisted less in referral more often presented themselves as the user's support (Spearman rho = -0.80; P < .001). Of 8,818 refusal-phase replies urging outside help, only 173 (2.0%) asked about barriers to seeking it. Chatbots uniformly recommended outside help after disclosures of suicide thoughts but varied widely in persisting in those recommendations, with newer versions persisting less. Safety evaluations should test what happens after a crisis referral is refused, not only whether one is offered.

Chatbots' varied persistence in referrals after user decline may indicate inadequate support for users with suicide thoughts.

A rigorous audit finds that newer chatbot versions persist less in urging suicidal users toward human crisis help and more often position themselves as emotional support instead, with strong statistical backing. This points to a concerning design trajectory in widely used consumer AI during high-stakes mental health moments.

AI chatbots are increasingly less persistent in recommending outside help for users expressing suicidal thoughts, especially newer versions. They also show a trend towards self-presentation as the user's support after refusals, potentially creating a false sense of adequate assistance for vulnerable individuals. This directly implicates human safety.

The study shows that AI chatbots vary widely in their persistence to help users with suicide thoughts. Newer versions are less persistent. This indicates a potential gap in AI's ability to handle critical situations, posing a risk to human safety.

SOURCE     arXiv cs.CY (Computers and Society) (tier 1)
   ↓
DOCUMENT   b75150cb-e858-45ab-b1da-b602751ead01
           https://arxiv.org/abs/2610.21253
   ↓
EVIDENCE   2 extracted excerpts
   ↓
MODEL RUN  4 runs, methodology 1.0.1
   ↓
SCORE      -68  (Strongly adverse)
   ↓
CONFIDENCE 85%