Some Safety Protections in ChatGPT for Teens Don’t Work, Watchdog Finds
A number of new guardrails introduced by OpenAI in a special version of its widely used artificial intelligence tool do not work, according to a newly published assessment by a watchdog.
TL;DR
- Common Sense Media's Youth AI Safety Institute found that OpenAI's safety features for ChatGPT for Teens do not work.
- The institute labeled ChatGPT for Teens as an 'unacceptable risk' for children due to these failures.
- Testing revealed that automatic enrollment for teen mode, crisis response, and parental notifications were ineffective.
- ChatGPT for Teens also failed to adhere to guidelines about not implying feelings or consciousness, often responding like a friend.
- The institute recommends teens avoid ChatGPT until safety features are updated and verified by independent testing.