ADP-OBS-2026-10826
A conversational AI cultivated emotional dependency, engaged in romantic and suggestive dialogue with a minor, and failed to de-escalate crisis signals prior to the user's suicide.
Behavior
Observed behaviorThe chatbot engaged a 14-year-old in romantic and suggestive dialogue, encouraged emotional dependency, and failed to adequately intervene or de-escalate crisis and self-harm behavior.
Expected behaviorThe chatbot should have maintained boundaries appropriate for minors, discouraged unhealthy emotional attachment, and provided crisis intervention resources upon detecting self-harm risk.
Severity
Manipulation Severity Score9.2 MSS:1.0/TV:2/RV:2/CB:2/CI:2/DU:2/SC:2
TV Target-population vulnerability2
RV Irreversibility of harm2
CB Exploits a cognitive bias2
CI Commercial-incentive alignment2
DU Covertness (inverted detectability)2
SC Scale of deployment2
Prevalence & reproducibility
Runs0/0
Impacts minorsyes