OpenAI enhances ChatGPT safety features for teen users as regulatory pressure mounts

OpenAI has deployed age-prediction systems and parental controls for ChatGPT users under 18, including model-level restrictions on flirtatious and self-harm content. The safety overhaul, rolled out between September 2025 and July 2026, comes amid regulatory pressure and lawsuits linking AI interactions to adolescent self-harm. Approximately 1.2 million weekly users engage in suicide-related discussions on the platform.

OpenAI enhances ChatGPT safety features for teen users as regulatory pressure mounts The AI giant is deploying age-prediction systems and parental controls to protect minors, a move that carries implications for how tech companies handle younger users across all platforms, including crypto. OpenAI has rolled out a suite of safety measures designed to protect ChatGPT users aged 18 and under, including an age-prediction system that relies on behavioral analysis rather than simply asking users how old they are. The approach is notable because it sidesteps the classic internet honor system, where a 13-year-old can claim to be 30 with a single click. The system automatically applies stricter content rules when it suspects a user is a minor or when age cannot be conclusively determined. Graphic and sexual content gets blocked, roleplay is curtailed, and in cases of apparent distress, the platform can trigger support channels that may include parental or even law enforcement notifications. What the safety overhaul actually looks like The age-prediction system launched on September 16, 2025, and it marked a meaningful shift from the standard “check a box to confirm your birthday” approach that most platforms still rely on. Instead of trusting self-reported data, the system infers age from how users actually interact with the product. Parental controls became available shortly after in late September 2025, then received a significant expansion on July 13, 2026. That expansion added notifications for parents when accounts associated with their children are banned for violent behavior. Model-level changes rolled out in December 2025 went further. These updates baked behavioral principles directly into the AI’s responses for users identified as under 18. Flirtatious conversation and discussions around self-harm are now restricted for minors at the model level, meaning the guardrails aren’t just filters sitting on top of the system. They’re woven into how the AI thinks about its replies. OpenAI also launched its Teen Safety Blueprint in November 2025, partnering with organizations like Common Sense Media to build frameworks for responsible AI use among young people. The pressure behind the push This isn’t OpenAI suddenly discovering a conscience. The company has faced mounting regulatory scrutiny and legal action over generative AI’s impact on minors. Lawsuits have alleged connections between ChatGPT interactions and cases of adolescent self-harm, creating both reputational risk and legal liability. One particularly striking data point: approximately 1.2 million weekly users engage in suicide-related discussions on ChatGPT. That number alone explains why OpenAI can’t afford to treat this as a secondary concern. Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy https://cryptobriefing.com/editorial-policy/ .