OpenAI expands ChatGPT for Teachers and publishes classroom usage data

OpenAI expanded ChatGPT for Teachers to 55 more US school systems across 20 states, adding more than 100,000 educators and staff and bringing the free programme to over 100 K-12 organisations across 30 states and more than 300,000 educators in total. The new agreements included a data-privacy framework spanning 16 states, which OpenAI said gives districts a common way to check the product against student-privacy requirements such as FERPA; data submitted through a school’s ChatGPT for Teachers workspace is not used to train OpenAI’s models by default. Access remains free for verified US K-12 educators through June 2028.

Alongside the expansion, OpenAI published a privacy-preserving analysis of how ChatGPT is used for learning, reporting that people of all ages have as many as 70 million conversations a week aimed at testing what they know — asking for a quiz, or checking a misconception. In the US specifically, messages related to classwork and homework peaked above 460 million a week during the school year, falling to above 180 million a week even over the summer.

The pairing followed OpenAI’s launch of ChatGPT for Teens eight days earlier, part of a broader back-to-school push into education that continued to draw scrutiny over student-data handling even as the company expanded free access.