Denver schools blocking students’ access to ChatGPT over concerns about group chats, adult content
Laurence Dutton via Getty Images
District officials said the popular AI chatbot’s new features could lead to cyberbullying and expose students to content related to self-harm and violence.
This article was originally published by Chalkbeat.
Denver Public Schools announced that starting this week, it is blocking students’ access to ChatGPT on school-issued devices and the district’s Wi-Fi network because of concerns about the artificial intelligence chatbot’s new features.
Those features include the ability for users to start a 20-person group chat and the planned addition of adult content to ChatGPT. In an email to families Friday, a top district official said DPS worries the new content could “potentially facilitate harmful interactions and/or the creation of content related to self-harm, violence and bullying.” The 20-person chat feature could lead to cyberbullying and academic misconduct, among other concerns, the email said.
Sam Altman, the CEO of OpenAI, the organization that created ChatGPT, said on X this past fall that the new version would allow “erotica for verified adults.”
DPS Chief Information Officer Richard Charles wrote in the email to families that it’s the district’s “responsibility to ensure students are only engaging with age-appropriate information and systems.”
Schools across the country have been grappling for years with how and when to use AI in the classroom. Some districts, including New York City Public Schools, initially blocked ChatGPT over concerns about student safety and learning before reversing course.
Denver Deputy Superintendent Tony Smith said DPS never explicitly approved the use of ChatGPT by its 89,000 students, and this latest announcement solidifies that position.
Instead, the district uses Google Gemini, which Smith said is more compatible with DPS’ monitoring system and data privacy rules. The district also uses MagicSchool, an education-specific AI tool created by a former Denver charter school principal that can help teachers create lessons or give students feedback on their writing.
“It’s not that we are scared of this technology or we’re trying to be behind the curve,” Smith said. “We’re trying to be strategic and thoughtful about the implementation of this technology.”
A district AI committee will meet later this month to discuss blocking ChatGPT for DPS staff as well, the email to families said.
One of the biggest pitfalls of AI, Smith said, is the potential for it to undermine students’ ability to think critically and express themselves in their own voice.
“We’re trying to make sure kids think and can access their skill sets and not ChatGPT,” he said.
Safety is another top concern. Smith said district officials have read about the ways in which AI can affect children’s mental health. Families in Colorado and other states have sued AI companies, alleging that their children committed suicide after developing relationships with chatbots.
DPS uses a digital tool called Lightspeed to monitor how students are using the internet, Smith said. But “we just can’t monitor everything,” he said. Given that, he said, district officials felt it was best to err on the side of safety.