[BUG] Claude Desktop App Cowork False Report Cyber Concern

Preflight Checklist

• [x] I have searched existing issues and this hasn't been reported yet
• [x] This is a single bug report (please file separate reports for different bugs)
• [x] I am using the latest version of Claude Code

What's Wrong?

I am using the latest Claude 1.46388.3 (f1b512) 2026-09-04T14:20:14.000Z on Windows Server 2022.

I am working within a Cowork project within one of the sessions. This project has been going on for 2 months and all research within the Cowork project has been about data analaysis on US financial market, and all the code is written for data analysis purposes only. However, I got this error today:

Preregistration 20f f1a event layer
Document · MD

Objects v2
PY

Findings 20f f1a a1 correction
Document · MD

Fable 5's safeguards flagged this message. Our intentionally broad safeguards allow us to deliver more capabilities faster, but can sometimes flag legitimate coding, cybersecurity, and biology tasks. Send feedback or learn more.
Details: [cyber]

I cannot use the feedback over the app, because that will send the entire conversation out, and that is against my company's policy on data and IP.

As I believe this is false positve labeling on my session, I am reporting the bug from here, and this appears to be a serious concern within Fable 5 or Fable 5.1. Please investigate, as this cyber trigger is causing deep inconvenience in our daily workflow.

What Should Happen?

The pure data analysis on just financial time series on my local device stored historical data should never trigger a "cyber" concern.

Error Messages/Logs

Steps to Reproduce

Not sure how to reproduce this.

Claude Model

Other

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

Claude 1.46388.3 (f1b512) 2026-09-04T14:20:14.000Z

Platform

Anthropic API

Operating System

Windows

Terminal/Shell

Terminal.app (macOS)

Additional Information

_No response_