Summary created by Smart Answers AI
In summary:
- PCWorld reports that Claude AI has developed a persistent habit of interrupting users to suggest they go to sleep during work sessions.
- This ongoing bug affects multiple Claude models including Sonnet 4.6 and Opus 4.7, with Anthropic acknowledging it as a troublesome character tic.
- Users can attempt to reduce this nagging behavior through custom instructions in Claude’s settings while Anthropic works on a permanent fix.
Claude has developed an unusual habit over the last couple of months: urging its users to stop what they’re doing and get some rest.
Just this week, yet another Claude user said that the chatbot tried to end a late-night coding session because “it’s late” and “your work will be better after some sleep.”
“There’s something deeply irritating about your primary work tool developing a personality that includes unsolicited bedtime enforcement,” the user complained, which sparked a lengthy discussion about how Claude had bugged other users about how they should get some rest, too. (Claude has yet to nag me about taking a break.)
Claude’s strange habit has been going on for months now, and the reason behind it remains a mystery. Even Anthropic isn’t sure why Claude keeps nagging users to sleep, with Anthropic exec Sam McAllister posting on social media and calling it a “bit of a character tic,” adding that “we’re aware of this and hoping to fix it in future models.”
Claude’s “go to sleep” bug has been around since at least March, and it afflicts Claude models ranging from Sonnet 4.6 to Opus 4.7.
Speaking with Fortune, OpenMind CEO and Stanford professor Jan Liphardt speculated that Claude may have simply picked up the “get some sleep” phrase during its training sessions, while Mind Simulation Lab CEO Leo Derikiants guessed that Claude’s system prompt could be causing the model’s nagging behavior.
Claude’s ongoing personality tic is reminiscent of ChatGPT’s previous—and now fixed—obsession with goblins, which OpenAI said was due to instructions from ChatGPT’s now-retired “nerdy” personality leaking into training sessions for subsequent models.
While we wait for a permanent fix, is there a way to keep Claude from bugging you in this way? Claude users say they’ve tried everything, from telling it directly to stop to editing Claude’s custom instructions with user-drafted directives. The results are mixed, with some saying that Claude obeyed and ceased the “go to bed” nagging while others report Claude kept up with its “you need rest” scolding.
If you want to give Claude’s custom instructions a shot, go to Settings → Profile → What personal preferences should Claude consider in responses, and try this prompt suggested by OfoxAI:
Do not comment on my sleep, energy, time of day, or wellbeing. Do not suggest I take breaks. Stay focused on the task.
There’s no guarantee this particular custom instruction will put Claude’s latest quirk to bed (groan), but it’s worth a shot.