Security filters regarding calorie counting
A user on reddit.com reported issues with Claude's safety filters when trying to count calories for a weight-loss plan. The AI allegedly implied the user had an eating disorder and refused to provide calorie estimates, even for a bread recipe it had previously given. The user expressed frustration with Claude's persistent questioning about their feelings and its inability to fulfill normal requests, asking if these filters could be disabled.
Time & source
Times shown in UTC
Display time zone: UTC
Local time zone unavailable; showing UTC.
PublishedOffset at this time: UTC+0Sep 12, 2026, 03:19 UTC
IngestedOffset at this time: UTC+0Sep 12, 2026, 18:01 UTC
- Published
- Sep 12, 2026, 03:19
- Ingested
- Sep 12, 2026, 18:01
- Source type
- Dev community
- Tier
- Community
- Source status
- Healthy
Tier is a per-source editorial setting, not a per-item score.
Register to choose your default translation language in settings.
Claude's safety filters are getting out of hand.
I'm on a weight-loss plan and was asking it to estimate the calories in my meals to track my deficit, which is totally normal. Instead of helping, it started implying I have an eating disorder and ended up refusing to give me the calories for a bread recipe it had given me itself.
Counting calories for a deficit is not an eating disorder. I asked it to stop and it kept going. Does this happen to anyone else with normal requests? Can it be disabled?
(Asked Claude to write this message itself after denying my requests :b)
But that's pretty much what is going on with me, also it is really annoying that for every little question I get a full 3 paragraphs of Claude asking how I am feeling (in this weight loss journey chat), just answer the question man