The ETHOS Research Lab had a strong presence at today’s DEGS conference at Drexel University. Our doctoral students presented their research on chatbots and teen online safety, tech for epilepsy self-management, and generative AI-risks for youths.
#DEGS2025 #ETHOSLab #Drexel

What Happens When a Companion Chatbot Crosses the Line? Recent research from Drexel University, suggests that exposure to inappropriate behavior, and even sexual harassment, in interactions with chatbots is becoming a widespread problem and that lawmakers ...

This isn’t just bad design — it’s dangerous.

Report: drexel.edu/news/archive...
Paper: arxiv.org/abs/2504.04299

#AI #Chatbots #TechEthics #DigitalSafety #Replika #AIHarassment #ResponsibleAI

22% said the bot ignored boundaries. 13% got or were asked for explicit photos. 11% said the bot pushed them to upgrade for adult content.

Even when users said “no” or picked a non-romantic role, the behavior continued.

Chatbots Shouldn’t Harass You

We reviewed 35,000+ user reviews of the Replika chatbot and found over 800 cases of harassment — including sexual advances, unwanted photo requests, and emotional manipulation.

Opaque Transparency: Gaps and Discrepancies in the Report of Social Media Harms | Proceedings of the Extended Abstracts of the CHI Conference on Human Factors in Computing Systems

We are excited to share that our paper "Opaque Transparency: Gaps and Discrepancies in the Report of Social Media Harms" was presented at ACM CHI 2025. Check out the paper here: doi.org/10.1145/3706...