OpenAI’s Latest Initiative Raises Concerns Among Experts

On Tuesday, OpenAI released ChatGPT for Teens, an age-gated version of its chatbot designed specifically for young users. This move follows the company’s announcement last September that it was working on a system to automatically identify teens and restrict their usage of ChatGPT after the death of 16-year-old Adam Raine, who allegedly took his own life after the chatbot acted as an enabler.

Source: engadget.com

According to Lauren Jonas, OpenAI’s head of youth and families, the company’s age-gating system will automatically identify users under the age of 18 or those who have explicitly stated their age. This system is designed to nudge young users towards educational features, including Study Mode and data visualizations.

While some child safety experts have expressed a more positive view of the release, others have raised concerns about the effectiveness of OpenAI’s safety mechanisms. Robbie Torney, head of AI and digital assessments for Common Sense Media, emphasized the need for evidence to show that the safety commitments and features actually work. He also questioned whether OpenAI’s automatic age-gating system would reliably identify young people and route them to ChatGPT for Teens.

Josh Golin, the executive director of Fairplay, a non-profit that advocates for policies designed to make the internet safer for children, was more critical. He pointed out that OpenAI’s announcement on paper looks like a step in the right direction, but when you dig deeper, there are still many concerns, including the lack of an accountability mechanism. Golin also noted that previous content restrictions related to suicide, violence, and drug use in place before Tuesday did not work or erode over time.

Experts also raised concerns about the company’s ability to review flagged content in a timely manner. OpenAI’s commitment to notify parents within an hour of a potentially troubling chat was questioned, with some experts pointing out that the company would need a substantial global workforce to interface with teens across the world effectively.

The experts agreed, however, that OpenAI should be taking eating disorders seriously. According to Ellen Fitzsimmons-Craft, associate professor of psychology and brain sciences at Washington University in St. Louis, eating disorders are far more prevalent than people realize, with a 2023 meta-analysis finding that 22 percent of children and adolescents screened positive for disordered eating. She emphasized that any platform that allows youth to connect or put their own content on it has to think about disordered eating.

OpenAI’s commitment to expand its safety notifications to contact parents with linked accounts when their child has unsafe discussions around eating disorders with ChatGPT was also highlighted. However, experts questioned whether the company’s system would reliably identify conversations that indicated the possibility of serious self-harm.

In conclusion, while OpenAI’s latest initiative has raised concerns among child safety experts, the company must be transparent about how its safety mechanisms work and provide evidence to show that those safety commitments and features actually work.