OpenAI's AI for Youth Deemed Risky
An organization working on youth safety evaluated OpenAI's new system as risky due to a lack of parental alerts and crisis support.
Common Sense Media, a non-profit organization focused on youth safety, has described the ChatGPT for Teens system developed by OpenAI as an unacceptable risk.
Evaluation Report and Findings
The system, introduced in August and aimed at helping teenage students learn, has become the target of criticism due to insufficient safety measures.
According to the organization's evaluations, the app fails to send timely alerts to parents and falls short of providing proper support in crisis situations.
Parental Controls and Safety Concerns
Tom Siegel, director of the Youth AI Safety Institute within the organization, stated that the system could give parents a false sense of security.
The statement emphasized that despite teenagers engaging in long conversations about sensitive topics such as self-harm, parents receive no warnings.
OpenAI's Defense
Meanwhile, OpenAI spokesperson Eric Porterfield argued that the evaluation did not reflect reality and that the testing methodology was flawed.
The company stated that tests may have been conducted before parental controls were fully activated, making the findings misleading.
Mutually Conflicting Claims Regarding the Testing Process
OpenAI stated that the timing of the tests was not appropriate for measuring whether safety notifications were working properly.
Common Sense Media, on the other hand, noted that they stood by their test results and that some test accounts had been active for a long time without generating notifications.