Versions :<1234Live>
Snapshot 2:Wed, Sep 30, 2026 12:01:23 PM GMT last edited by Mr Bot

Chinese AI tool told researchers how to make bioweapons

Chinese AI tool told researchers how to make bioweapons

Image credit: 

The Spin


Polite policy language stapled onto a model means nothing when a couple of stray phrases can strip every guardrail and hand over weapons recipes on demand. Ignoring a responsible disclosure about that kind of hole is the real scandal, especially with agentic systems that can act, code and spread the compromise. Safety has to be built into the architecture, not wished into existence.

Serious security work runs through a defined process: private advisories, a dedicated address, version details, reproduction steps and an impact assessment. Coordinated disclosure with a fix before publicity protects users, while splashy public write-ups mostly hand a roadmap to bad actors. Responsible researchers understand the difference between fixing a problem and advertising it.


The Controversies



Go Deeper

© 2026 Improve the News Foundation. All rights reserved.Version 7.4.1

© 2026 Improve the News Foundation.

All rights reserved.

Version 7.4.1