AFT President Calls for Cautious AI Integration in Schools

Randi Weingarten, president of the American Federation of Teachers, warned that students must master critical thinking before relying on generative artificial intelligence. She emphasized that human judgment, not machines, should guide AI use in education and highlighted two primary worries: student privacy and “cognitive offloading,” where learners depend on AI to complete assignments.

The Hill

When AI tools first entered the market, many districts responded with outright bans. Over time, those prohibitions have softened as educators recognize the technology’s relevance to the modern workplace. Schools are now experimenting with AI-detection software to identify unauthorized use and have begun partnerships—such as a safety and privacy agreement with Microsoft—to assure parents that any AI deployment follows strict safeguards.

The Hill

Weingarten argued that classrooms must remain “safe and welcoming” while engaging students in active learning. She called for federal leadership on AI safety standards, suggesting that former first-ladies should champion such efforts. Her remarks position the teachers’ union as a voice urging policymakers to prioritize human-centered education before expanding AI tools.

The Hill

• “So, you know, from an education and a people perspective, the human beings have to be in charge of AI. The robots can’t be in charge. The machines can’t be in charge. And AI as a generative technology cannot outstrip human decision-making,” — Randi Weingarten

The Hill

• “What I’m saying is that we need to make learning engaging. We need classrooms to be safe and welcoming and we need learning to be engaging. And one of the things we just did was we just did a safety and privacy agreement with Microsoft so that we could actually assure parents throughout the country that if that if AI is being used, it’s being used safely,” — Randi Weingarten

The Hill

If schools adopt AI without clear guidelines, the risk of reduced critical-thinking skills and compromised student data could rise. Conversely, a measured approach—combining robust privacy agreements, detection mechanisms, and curriculum that foregrounds thinking—could help students become proficient users of AI rather than passive consumers. The union’s stance signals a push for national policy that balances technological benefits with educational integrity.

The Hill