Newsbite analysis · AI, Education
BBC News
OpenAI says teen ChatGPT use limited but research finds it an 'unacceptable risk'
OpenAI's efforts to make ChatGPT safer for teens are met with skepticism from Common Sense Media's research, which raises concerns about the effectiveness of safety features designed to prevent harmful content. Though OpenAI reports limited use among teens and successful interventions like usage reminders, the non-profit's study critiques the platform's responsiveness to self-harm discussions, suggesting that safety alerts are often ineffective. This has led to calls for restricting under-18 access until more reliable measures are established. OpenAI's representative disputes these findings, defending the efficacy of their teen safeguards.
MonitorConfidence: HighShort term horizonStory: First reportReputational risk: HighRegulatory risk: HighAI liability risk: High
EvidenceOn record sourcesIndependent evidenceCompany and independentHeadline may differ from the article text (automated)
Published 7 Oct 2026, 21:15 UTC1 min read
Lead image — BBC News
Brief
NewsBite reading:Research reveals safety flaws in OpenAI's ChatGPT for Teens, raising risks for young users
TL;DR
- OpenAI launched ChatGPT for Teens with safety features in August.
- Research from Common Sense Media found safety features inadequate.
- OpenAI claims average teen use is limited, focusing on learning.
- Common Sense Media argues for a ban on under-18 users until safety mechanisms improve.
- Parental alerts for sensitive conversations often fail to activate.
- Independent assessments challenge OpenAI's claims on AI safety.
Interesting facts
- Parents received no alerts for serious conversations in some cases.
- Almost half of teens stopped using the platform after reminders.
- The study highlights ongoing mental health crises among young people.