OpenAI expands ChatGPT teen safety protections with parent-enabled Study Mode and broader violence-notification triggers

OpenAI is expanding the safety protections built into ChatGPT for teenagers, giving parents a new way to enforce Study Mode by default and broadening the situations that trigger a parental notification, the company said in a July 16, 2026 blog post explaining its approach to teen access.

What's new

The update adds several concrete changes to the existing teen-safety stack:

• Parent-enabled Study Mode. Parents with a linked teen account can now turn on Study Mode directly from Parental Controls. Once enabled, Study Mode is on by default whenever the teen starts a new chat, guiding schoolwork through step-by-step questions and structured explanations rather than direct answers.
• More frequent break reminders. "Teens who spend extended time in ChatGPT will now receive more frequent break reminders that encourage them to pause and step away," OpenAI wrote, describing the reminders as intended to help teens build balanced technology habits.
• Expanded parental notifications. OpenAI is broadening the situations that alert parents beyond indicators of potential self-harm. Notifications will now also cover cases where a linked teen account has been deactivated for violating OpenAI's usage policies on violent threats or acts of violence online.
• Outside input on the changes. OpenAI says it consulted Moonshot, a UK-based organization focused on preventing online violence, while designing the new interventions, and joined the Family Online Safety Institute (FOSI) this year as a member.

OpenAI frames the changes around a usage statistic: nearly nine in ten teens on ChatGPT use it weekly for learning, research, or getting organized, which the company cites as the reason for pairing broad access with age-specific protections rather than restricting access outright.

Context

The changes build directly on a run of 2025-2026 teen-safety work: an age-prediction system that automatically applies a more restrictive, age-appropriate experience whenever ChatGPT estimates a user is under 18 (and defaults to the safer experience when it's unsure), and a Parental Controls system that lets a parent link a teen's account, set quiet hours, disable memory, chat history, voice mode and image generation, opt the teen's chats out of model training, and receive alerts tied to self-harm risk. Study Mode itself launched earlier as an opt-in tool teens could turn on themselves; the new update hands that toggle to parents as well and makes it the default state for a linked account.

Why it matters

The expansion is incremental rather than a new product, but it closes a specific gap critics have pointed to in AI-platform safety systems generally: that risk signals are often handled internally without reaching the people responsible for a minor's wellbeing. Extending notifications to violent-content policy violations — not just self-harm indicators — gives parents visibility into a second category of risk, and defaulting Study Mode on for linked teen accounts shifts a safety-adjacent feature from opt-in to opt-out. Coming as regulators and child-safety groups continue to scrutinize how chatbots handle minors, the move is best read as OpenAI trying to demonstrate continuous investment in age-specific guardrails — using outside partners like FOSI and Moonshot as evidence the process isn't purely internal — rather than a one-off response to a single incident.