Parental controls on five major AI assistants (ChatGPT, Gemini, Meta AI, Siri AI and Copilot) turn out to be thinner than the safety messaging suggests. A hands-on comparison found that most depend on a teenager agreeing to be supervised.
The comparison covered five of the seven chatbots its author analyzed. Age rules differ. ChatGPT, Meta AI, Siri AI and Copilot set a minimum age of 13 for their accounts. ChatGPT requires parental consent under 18 and predicts age from signals such as time of day and topics. Gemini follows Google account rules: no under-13s unless the account sits in an adult-managed family group. Apple lets an adult create an account for a younger child, but that account only reaches the older Siri.
Oversight varies more. ChatGPT lets parents set time limits and adjust sensitivity, but not view conversations, and teens can break the link. Meta's Family Center gives parents topic summaries and self-harm alerts, without transcripts. Microsoft's Family Safety portal can switch Copilot on or off, set hours and disable features like image generation, but parents cannot read chats. Gemini offers only an on/off switch. Apple's new iOS 27 controls are optional for teens.
Two patterns stand out. For ChatGPT, Meta AI, Siri AI and Copilot, the source says linking a parent account is optional or revocable for teens. And ChatGPT and Gemini both allow anonymous browser use. OpenAI says that route gets the default under-18 limits, while Gemini's anonymous chats get only the general rules, not the minor-specific ones. Google also admits its filters are not perfect.
The design logic resembles social media's earlier teen-account rollouts: protections built around the teen's cooperation. That sits awkwardly beside governments age-gating more services. The open question is whether regulators will accept opt-in supervision, or push for controls a teenager cannot switch off.