OpenAI launches ChatGPT version for teens with stricter safeguards

OpenAI's launch of a teen-focused ChatGPT version directly addresses a growing legal and reputational crisis. Multiple families have brought suits against the company after their children encountered harmful content on the platform, with at least two cases resulting in serious injury or death. Florida's decision to file the first state-level suit in June signals a broader regulatory shift beyond individual litigation, indicating that policymakers and attorneys general are viewing child safety on AI platforms as a priority concern.

The product's design reflects a tension inherent in moderating AI for minors: OpenAI chose not to rely on explicit age verification, instead deploying a behavioral-detection system tuned to over 2,000 signals. This approach avoids friction (no login barriers or ID requirements) but depends on pattern-matching accuracy. The homework feature—substituting guided questions for direct answers—aligns with educational principles of active learning, though it remains to be seen whether teenagers will experience it as genuinely helpful or merely evasive. The emotional-claim restriction is notably granular: it targets a specific anthropomorphization risk without blocking the conversational interface entirely.