OpenAI’s ChatGPT for Teens Faces Expert Skepticism Over Safety Mechanisms

OpenAI’s recent rollout of ChatGPT for Kids, featuring age-based restrictions, has drawn sharp criticism from child safety advocates who question the transparency and efficacy of its safeguards.
Safety experts emphasize that independent testing is essential before the platform is recommended to families. Robbie Tynjala of Common Sense Media stressed that OpenAI must provide concrete proof that its safety features actually work, rather than relying on unverified claims.
Josh Golin, head of the Fairplay initiative, raised serious concerns about the lack of accountability. He noted that OpenAI’s past restrictions on sensitive topics like suicide and violence have often failed, arguing that self-regulation is insufficient without robust oversight.
Experts also doubt the reliability of the automatic age-verification system. Tynjala pointed out that OpenAI has not released data on false positive or negative rates, making it clear that automated systems cannot perfectly identify young users. Additionally, teenagers are likely to find ways to bypass these checks.
Skepticism remains high regarding OpenAI’s promise that flagged content will be reviewed by staff and parents notified within an hour. Tynjala questioned the trustworthiness of the automated filters, while Golin demanded hard data on daily flag volumes, team size, and review times, noting it is unrealistic to expect sufficient staffing for thorough manual review of every dangerous interaction.
Finally, experts argue that requiring parents to manually link accounts undermines protection. Golin suggested that safety features should be built into the product by default, such as automatic time limits, rather than placing the burden of complex configuration entirely on parents.
Related article
Fortinet’s Virtue AI Deal Bolsters Agentic AI Security
Ken Xie, Founder and CEO of Fortinet | Credit: FortinetFortinet acquires Virtue AI to strengthen runtime protection and automated validation for AI systems as enterprises scale deployment of agents and modelsFortinet has acquired Virtue AI, a company
xAI Unveils Early Grok Beta for Complex Programming Tasks
On May 14, Elon Musk’s AI venture xAI launched an early beta of a coding assistant named Grok Build. Currently restricted to SuperGrok subscribers, the platform is designed to streamline software engineering and handle intricate programming workflows
Anthropic CEO Outlines Plan To Pace The Frontier
AI researchers have issued increasingly urgent warnings about artificial intelligence risks, with OpenAI CEO Sam Altman suggesting it may be time to “pace” development. But what would that actually entail?In a recent blog post, Anthropic CEO Dario Am
Related Special Topic Recommendations
Comments (0)
0/500

OpenAI’s recent rollout of ChatGPT for Kids, featuring age-based restrictions, has drawn sharp criticism from child safety advocates who question the transparency and efficacy of its safeguards.
Safety experts emphasize that independent testing is essential before the platform is recommended to families. Robbie Tynjala of Common Sense Media stressed that OpenAI must provide concrete proof that its safety features actually work, rather than relying on unverified claims.
Josh Golin, head of the Fairplay initiative, raised serious concerns about the lack of accountability. He noted that OpenAI’s past restrictions on sensitive topics like suicide and violence have often failed, arguing that self-regulation is insufficient without robust oversight.
Experts also doubt the reliability of the automatic age-verification system. Tynjala pointed out that OpenAI has not released data on false positive or negative rates, making it clear that automated systems cannot perfectly identify young users. Additionally, teenagers are likely to find ways to bypass these checks.
Skepticism remains high regarding OpenAI’s promise that flagged content will be reviewed by staff and parents notified within an hour. Tynjala questioned the trustworthiness of the automated filters, while Golin demanded hard data on daily flag volumes, team size, and review times, noting it is unrealistic to expect sufficient staffing for thorough manual review of every dangerous interaction.
Finally, experts argue that requiring parents to manually link accounts undermines protection. Golin suggested that safety features should be built into the product by default, such as automatic time limits, rather than placing the burden of complex configuration entirely on parents.
Fortinet’s Virtue AI Deal Bolsters Agentic AI Security
Ken Xie, Founder and CEO of Fortinet | Credit: FortinetFortinet acquires Virtue AI to strengthen runtime protection and automated validation for AI systems as enterprises scale deployment of agents and modelsFortinet has acquired Virtue AI, a company
xAI Unveils Early Grok Beta for Complex Programming Tasks
On May 14, Elon Musk’s AI venture xAI launched an early beta of a coding assistant named Grok Build. Currently restricted to SuperGrok subscribers, the platform is designed to streamline software engineering and handle intricate programming workflows
Anthropic CEO Outlines Plan To Pace The Frontier
AI researchers have issued increasingly urgent warnings about artificial intelligence risks, with OpenAI CEO Sam Altman suggesting it may be time to “pace” development. But what would that actually entail?In a recent blog post, Anthropic CEO Dario Am





Home






