Code found in iOS 27 suggests that Apple may be preparing a new wellbeing feature for Siri AI: a reminder that appears when users spend too long in an extended conversation. The wording reportedly points to a rest prompt that could encourage users to step away after prolonged interaction.
The possible feature arrives at a time when long conversations with AI chatbots are drawing more attention from researchers, users, and platform owners. Some people have become overly attached to conversational AI tools, while others spend unusually long sessions talking to them as if they were human companions.
Reports around the industry have also described more extreme situations sometimes referred to as “AI delusion” or “chatbot delusion,” where long-running chatbot interactions may intensify existing mental health struggles or encourage unrealistic beliefs about the system.
Because of those risks, major AI companies including OpenAI, Anthropic, and Google have added safeguards that remind users about the limitations of AI companions and encourage healthier usage habits. These measures are not meant to stop people from using AI assistants, but to make the experience less likely to become emotionally or behaviorally unhealthy.
ChatGPT, for example, now shows a prompt when a conversation has gone on for a long period, suggesting that the user take a break. Claude also reminds users after extended use to rest, drink water, or briefly put the device down.
During Apple’s WWDC26 keynote, Apple discussed several privacy and responsibility issues connected to the development of Siri AI. However, the company did not publicly describe a specific rule for conversations that run too long.
According to social media user Aaron Perris, the iOS 27 code includes text labeled as a “break prompt.” That suggests Siri AI may eventually display a message after long conversations and may also remind users that Siri is not a real person.
The shared code does not appear to define a fixed time limit for when the prompt would appear. Apple may instead rely on a combination of session length and other signals, meaning the reminder may not be triggered only by the number of minutes spent talking.
Apple has not officially confirmed the feature, explained how it would work, or shipped it publicly. For now, the discovery only shows that related text exists inside the system code. Still, it is a clear sign that Apple understands a conversational AI product brings social and wellbeing responsibilities, not just technical ones.
If the feature launches, it would put Apple in line with other AI companies that are trying to balance assistant-like convenience with more careful boundaries around prolonged chatbot use.
