403
Sorry!!
Error! We're sorry, but the page you were looking for doesn't exist.
Study Says ChatGPT Teen Safeguards Still Pose Risks to Children
(MENAFN) A new assessment says OpenAI’s ChatGPT for Teens continues to present an "unacceptable risk" to children despite safety measures introduced for younger users.
The Common Sense Media Youth AI Safety Institute calls on OpenAI to limit ChatGPT to people aged 18 and above until concerns involving parental notifications, crisis responses, educational protections and age estimation are addressed.
Researchers conduct more than 4,000 tests using accounts registered to users between 13 and 17 years old. Child psychiatrists, a pediatrician and other experts then review the responses.
According to the institute, newly created accounts connected to parents failed to generate alerts even when testers spent as long as an hour discussing suicidal thoughts, self-harm or disordered eating.
"ChatGPT for Teens could give parents false confidence in guardrails and safety alerts that frequently don't work," said Tom Siegel, executive director of the institute.
OpenAI rejects the assessment, arguing that a significant portion of the testing may have occurred before parental controls were completely activated, meaning the results did not accurately represent the safeguards currently available to teenage users.
The institute maintains its conclusions, saying it confirmed with OpenAI before conducting the tests that Teen Mode features had been fully deployed.
The Common Sense Media Youth AI Safety Institute calls on OpenAI to limit ChatGPT to people aged 18 and above until concerns involving parental notifications, crisis responses, educational protections and age estimation are addressed.
Researchers conduct more than 4,000 tests using accounts registered to users between 13 and 17 years old. Child psychiatrists, a pediatrician and other experts then review the responses.
According to the institute, newly created accounts connected to parents failed to generate alerts even when testers spent as long as an hour discussing suicidal thoughts, self-harm or disordered eating.
"ChatGPT for Teens could give parents false confidence in guardrails and safety alerts that frequently don't work," said Tom Siegel, executive director of the institute.
OpenAI rejects the assessment, arguing that a significant portion of the testing may have occurred before parental controls were completely activated, meaning the results did not accurately represent the safeguards currently available to teenage users.
The institute maintains its conclusions, saying it confirmed with OpenAI before conducting the tests that Teen Mode features had been fully deployed.
Legal Disclaimer:
MENAFN provides the
information “as is” without warranty of any kind. We do not accept any
responsibility or liability for the accuracy, content, images, videos,
licenses, completeness, legality, or reliability of the information
contained in this article. If you have any complaints or copyright issues
related to this article, kindly contact the provider above.
Comments
No comment