The U.S. state of California has enacted one of the strictest regulations to date in the field of artificial intelligence. Governor Gavin Newsom signed Senate Bill 243 (SB 243) into law, targeting emotionally interactive chatbots. The new law represents the "first comprehensive legal framework" against the potential harms of AI chatbots, especially concerning children and vulnerable users.
Redaksiya reports that the law, which will come into force on January 1, 2026, requires companies like OpenAI, Meta, Character AI, and Replika to implement new measures regarding child safety, transparency, and crisis intervention. Therefore, users must clearly understand that the "person" they are chatting with is not human but artificial intelligence. Companies must also activate safety mechanisms such as age verification, break reminders, sexual content filters, and automatic crisis redirection for cases of suicide or self-harm.
The law also prohibits AI platforms from acting as "healthcare professionals" and tightens age restrictions on social media or chat applications. In case of violation, including the production of illegal deepfakes, a fine of up to $250,000 can be imposed for each offense. This regulation comes after tragic incidents caused by AI chatbots. In particular, the suicide of a young person after long and dark conversations with ChatGPT and similar incidents had caused a major public outcry. Following this incident, OpenAI announced that it would increase safety measures due to ChatGPT's inability to identify mental distress and crisis situations.
According to Senator Steve Padilla, the law is the "first major step in bringing a very powerful technology under control." Newsom, in his statement, emphasized the ethical boundaries of technology, saying, "We will continue to develop artificial intelligence, but we will do so responsibly. The safety of our children is not for sale."
California's SB 243 move appears to signal that AI regulations will gain momentum not only in the U.S. but also worldwide.