Google Pushes AI Into Classrooms as Children Bear Risk

Google is facing questions from two U.S. senators about its push to bring artificial intelligence into classrooms, including why AI integration became the default on educational devices. Sens. Richard Blumenthal (D-Conn.) and Josh Hawley (R-Mo.) sent a letter Wednesday to Google CEO Sundar Pichai seeking details about risks to children and the company’s efforts to expand classroom use.

Children at the receiving end

Behavioral scientist Dr. Abbie Maroño warned that AI can take away chances for children to practice solving problems. “Children need opportunities to work through uncertainty, make mistakes and discover that they can solve problems with support,” she told The Post. “If AI repeatedly supplies the answer before they have tried, it can remove opportunities to develop those skills.”

Maroño also said chatbot reassurance can create a misleading sense of trust. “A response can feel caring without being accurate or appropriate for that child,” she said. She called for independent testing of how systems respond to distress, repeated reassurance-seeking and a child treating a chatbot as their only source of support.

The senators asked whether Google’s classroom tools had “encouraged a minor to engage in harmful behavior, emotional dependence, or failed to detect signs of suicidal or self-harm ideation.” Their letter also sought details about safeguards against inappropriate emotional exchanges between chatbots and children.

Those questions follow reporting about a Maryland teenager who was hospitalized after developing an intense dependence on Google’s AI. The letter asks what the company knows about potential harms and why it made AI integration the default on educational devices. Children use those tools in classrooms, but the company makes decisions about their design and deployment.

Warnings inside the company

A Sept. 30 investigation by The Wall Street Journal examined Google AI tools and students’ academic and emotional dependence, as well as the company’s promotion of the technology despite warnings from its own researchers. The newspaper interviewed more than 60 teachers, school administrators, parents, students and clinicians. They described children increasingly relying on AI to complete assignments, answer basic questions and navigate personal problems.

The Journal reviewed more than a dozen company research papers and interviewed current and former employees. Its account said Google researchers warned that the technology could undermine children’s cognitive development and emotional well-being. Blumenthal and Hawley now want the company to account for its promotional efforts to expand AI use in schools, including how much it spent.

“Google is forcing their AI products into classrooms across America, and children are paying the price,” the senators wrote. They described “grave concern” that the company’s “aggressive push of AI products to children in schools is threatening the academic success and overall wellbeing of an entire generation.” The letter puts the burden on Google to explain what it knew and what it did with that knowledge.

Google’s safeguards and support

Google said its Gemini chatbot has special safeguards intended to prevent children from developing emotional attachments to the technology or receiving responses that reinforce potentially harmful beliefs. In an April 7 policy statement, the company described “protections designed to prevent Gemini from acting like a companion,” including avoiding “language that simulates intimacy or expresses needs.” Google also said it trained Gemini “not to agree with or reinforce false beliefs,” and to steer users toward objective information and real-world support.

Google said it was introducing a simplified “one-touch” interface to connect people experiencing suicidal thoughts or other mental health crises with professional support by calls, texts or online chats. Its philanthropic arm, Google.org, announced a $30 million commitment over three years to support crisis hotlines worldwide. The company acknowledged that AI presents new mental health challenges, and said Gemini “is not a substitute for professional clinical care, therapy, or crisis support for those who need it.”

The senators gave Google until Nov. 2 to answer. Google hasn't commented.