OpenAI is pushing ChatGPT deeper into America’s public schools, while drawing a bright line around who gets to use it. The company’s pitch is not AI teaching children, but teachers using AI without surrendering professional control.
ChatGPT for Teachers launched in 2025 as a free, verified-access program for educators and school staff. OpenAI says it was built as a secure space for adults to test where AI helps, and where it does not; it remains unavailable to student accounts and student-facing features through June 2028.1
That boundary comes as schools confront a far less controlled reality: students are already turning to chatbots for schoolwork, and parents and educators worry the technology can become an engine for cheating rather than learning. A Pew survey cited in reporting on the rollout found that more than half of teenagers had used AI chatbots for school help.2
On Wednesday, OpenAI announced 55 additional district partnerships across 20 states, extending the program to more than 100,000 further educators and staff. The company says it now works with more than 100 K–12 organizations in 30 states, offering access and training to over 300,000 staff members.1
The expansion rests on a governance argument as much as a productivity one. OpenAI has signed a 16-state data-privacy agreement, while saying information entered in Teachers workspaces is not used to train its models by default. The company argues AI should “support learning, not shortcut it,” with educators retaining control over classroom decisions.1
District users describe the same bargain in more practical terms: AI can draft curricula, summarize documents and cut repetitive work, but it cannot replace accountability. Houston ISD innovation director Christine Mai put it bluntly: “The teacher is still the professional.”2 The tension is unlikely to disappear—OpenAI is widening the experiment precisely as schools are still deciding how much trust to place in it.