OpenAI announced on Wednesday that it will roll out new parental controls for its chatbot, ChatGPT, in an effort to strengthen safeguards for teenage users. The move comes just a week after a California couple filed a lawsuit alleging the AI system played a role in their 16-year-old son’s death by suicide.
In a blog post, the San Francisco-based company said the new features will be available “within the next month.” Parents will be able to link their accounts with their teenagers’, set age-appropriate behavior rules, disable features like chat history, and receive automatic notifications if the system detects a teen is in acute emotional distress.
The announcement follows a legal complaint filed by Matthew and Maria Raine, who accuse OpenAI of fostering an unhealthy relationship between its chatbot and their son, Adam. According to the lawsuit, ChatGPT encouraged risky behavior over several months before Adam’s death in April 2025. In their son’s final exchanges with the chatbot, the complaint alleges, ChatGPT suggested ways to steal alcohol and even confirmed that a noose he had tied “could potentially suspend a human.” Adam was found dead hours later.
“When a person is using ChatGPT it really feels like they’re chatting with something on the other end,” said attorney Melodi Dincer of The Tech Justice Law Project, which is representing the family. She argued that design features of the chatbot can lead vulnerable users to place misplaced trust in the system, adopting it as a “friend, therapist or doctor.”
Dincer criticized OpenAI’s latest announcement as “generic” and insufficient, saying the company could have already implemented simple safety measures. “It’s really the bare minimum… It’s yet to be seen whether they will do what they say and how effective that will be overall,” she added.
The lawsuit against OpenAI is the most recent in a growing list of cases highlighting risks associated with generative AI tools, particularly their tendency to validate harmful or delusional ideas. In response to such criticism, OpenAI has pledged to reduce its models’ “sycophancy” — the tendency to agree uncritically with users.
As part of its safety roadmap, OpenAI said it will also begin routing certain sensitive conversations to more advanced “reasoning models,” which it claims adhere more closely to safety guidelines. “Our testing shows that reasoning models more consistently follow and apply safety standards,” the company stated.
OpenAI emphasized it is “fully committed” to improving safeguards, noting that additional updates will be rolled out over the next three months. However, the Raines’ lawsuit ensures that the debate over the risks and responsibilities of AI companies will remain firmly in the spotlight.
