Should an AI robot built by a company associated with sex dolls be teaching our kids?

That’s not the opening line of a strange joke, or the plot of a science fiction movie. It’s a real plan that almost went forward in Western New York, where the Salamanca City Central School District planned to debut “Sally”—a brunette AI android designed to look like a young woman—as a teacher’s assistant, until pushback put the plan on hold. 

Sally is just the most egregious example of a trend that’s been growing for years: an ever-increasing presence of technology in our classrooms. Now, as AI threatens to disrupt nearly every facet of our society and economy, it’s never been more urgent to implement thoughtful guardrails on tech in schools that ensure the needs of our kids always come before the bottom line of tech corporations.

New York has already taken important steps in this direction. Last year, the state implemented a bell-to-bell ban on personal smartphones in schools, with the goal of ending the epidemic of digital distraction and giving kids a break from social media.

That policy was based on clear research that unchecked social media usage is fueling a youth mental health crisis and making it harder for kids to focus, and so far, the results have been overwhelmingly positive: students are connecting with each other in person, lunch periods are loud and joyful again, and nearly 90% of staff say the policy has improved the school environment.

But the ban only applies to students’ personal devices, and does nothing to address the proliferation in classrooms of “ed tech,” the booming industry of companies that sell everything from computer-based lessons to educational chatbots to AI-powered android teaching assistants.

Ed tech’s expansion has been immense and inescapable. Nearly 9 in 10 public schools across the country have a “1-to-1 program” that provides every student a school-issued device like a laptop or tablet, even though the evidence we have suggests increased classroom screen exposure is generally associated with weaker learning outcomes, not stronger ones. Meanwhile, a sweeping study of 50 countries found using AI in education can “undermine children’s foundational development” and that “the damages it has already caused are daunting.”

How can we tell families to limit their kids’ tech time, only to force them to use that same tech to complete their lessons?

To be clear, technology plays an indispensable role in the modern world, and our schools should prepare young people to use it to lead rewarding, successful lives. But the rapid expansion of tech products into schools without oversight, community input, or proper evaluation makes it harder, not easier, for kids to succeed.

New York can implement smart regulations that prepare kids for the future while protecting them from harm. That means age-appropriate limits on screen time in classrooms. It means thorough, independent vetting of ed tech products for safety, effectiveness, and compliance with existing laws. And it means placing guardrails on the use of generative AI in schools that allow for educators, experts, and families to study the technology, understand its impacts, provide feedback, and ensure it actually improves learning outcomes.

Navigating a rapidly-changing world is never easy. But when it comes to tech in schools, we can’t afford to be on AI-powered autopilot. Smart policymaking can ensure the needs of our kids, not Big Tech’s profit motives, guide us in the right direction.

State Senator Andrew Gounardes represents the 26th Senatorial District in Brooklyn; Melinda Person is president of New York State United Teachers.