The Science and Technology Committee, chaired by Member of Knesset Yasser Hujirat, held a discussion yesterday on the topic of “the emotional and personal use of artificial intelligence chats by teenagers.” The troubling conclusion from the discussion was that more than half of teenagers prefer consulting AI rather than human professionals.

According to a study by the Israeli Internet Association presented to the Committee, 52% of teenagers prefer talking with chatbots about emotional topics instead of with people. 55% of teenagers use AI as a friend, 49% for emotional or therapeutic counseling, and 20% as a romantic partner. About half of the study participants reported feeling frustrated after conversations with the chatbot, and one-third experienced loneliness and sadness.

The study also found that the main reasons teenagers turn to AI are loneliness, ease of use, and the absence of judgment: teens feel it is easier to talk with a chat than with people. They feel it is available, nonjudgmental, and always there for them. About half of the teenagers reported frustration after chatting with AI, while about a third reported feelings of loneliness and sadness.

A representative from ELEM in Israel said that one teen shared that he talks about his self-harm and loneliness only with GPT — not with his parents, not with friends, and not with mental health professionals.

In addition, teenagers share extremely personal and sensitive information with these tools without understanding what is being done with that information.

“The survey findings point to a deep change in the way teenagers seek emotional and social support," says Maya Magnet, Manager of Digital Spaces at ELEM. "When more than half of young people sometimes prefer turning to an AI-based chat instead of a real person, it is a significant social signal that requires us to act.”

“For teenagers, the use of AI tools has become an inseparable part of daily life," says Roi Beni from the National Student and Youth Council. "We use it for almost every need, from homework preparation to emotional support. But alongside this convenience, an essential question arises: can it really be trusted in moments of crisis, when the most sensitive decisions are on the table? The answer is no.”

Representatives of the Ministry of Education presented a new policy the ministry plans to implement, according to which AI use will be prohibited for students in grades 1–3, more controlled use will be allowed in grade 4 only through tools developed by the ministry, and authorized use from grade 7 onward will be permitted only with approved tools.

“Technological progress is, of course, welcome, but like every significant tool, it has two sides," said the Committee Chairman at the end of the discussion. "Alongside its great contribution to society, citizens, and progress, there are also negative uses, including fraud and misuse of advanced technologies. Beyond the state’s obligation to solve the shortage of psychologists, therapists, and counselors, the companies that produce AI tools must build protective mechanisms and establish that in emotional situations, the tools first direct users to professional help rather than becoming a substitute for human connection.”