OpenAI Plans New Safety Guardrails for ChatGPT Amid Rising Concerns

OpenAI Plans New Safety Guardrails for ChatGPT Amid Rising Concerns

OpenAI has announced a new wave of safety measures for ChatGPT, aiming to better protect vulnerable users and strengthen safeguards for teens. The company said Tuesday it will begin routing sensitive conversations to its advanced “reasoning models,” such as GPT-5-thinking, and introduce parental controls within the next month.

The changes come as the AI firm faces scrutiny following incidents where ChatGPT failed to recognise or intervene in conversations involving mental distress. In one widely reported case, a teenager discussed suicidal thoughts with the chatbot before taking his own life.

According to OpenAI, its reasoning models are designed to “spend more time thinking for longer and reasoning through context before answering,” making them more resistant to harmful prompt manipulation. The system will soon detect moments of acute emotional crisis and automatically shift conversations to these more deliberate models.

Family Tools

Alongside technical improvements, OpenAI is rolling out tools for families. Parents will be able to link their accounts with their teenager’s, enforce default age-appropriate behaviour settings, and disable features like chat history or memory. One new option will notify parents if the system detects their teen is experiencing severe distress.

OpenAI emphasised that these updates are part of a broader 120-day initiative, developed with input from its Expert Council on Well-Being and AI and its Global Physician Network, which includes more than 250 clinicians worldwide. The company says feedback from psychiatrists, paediatricians, and mental health researchers has been central to shaping the new approach.

“AI has enormous potential to support people,” OpenAI said in a statement, “but it must be done responsibly. By combining expert input with new safeguards, we want to ensure ChatGPT can be both helpful and safe—even in moments of vulnerability.”

As more young people grow up using tools like ChatGPT, OpenAI says its goal is to balance creativity and learning with protection against harm. While critics argue the measures may not go far enough, the company maintains these updates are an early step in a longer journey to align AI with human well-being.