Anthropic's Project Glasswing Update

When the model finds a vulnerability, it also finds a fix. Anthropic only shares the vulnerability with the Open Source maintainer, not the fix. Paying customers get fixes, confirming that the model does generate fixes for the vulnerabilities.

Sharing the vulnerabilities but not the fixes does sound like a shakedown operation.

Re: Anthropic's Project Glasswing Update

#5

Post by jerrythegerbil » Mon, Jun 08, 2026, 1:14 PM UTC

Glasswing and Mythos are wildly impressive.

The team writing about it has a core charter to publish research about how AI will be disruptive to certain industries. The publication of such research is the disruption.

What remains when you stop gamifying the lag time of putting onus of counter evidence of impact and not just minmaxxing the discovery of bugs at the start of a development process is…

Does anyone remember LK-99? Yeah. Playbook works.

Re: Anthropic's Project Glasswing Update

#6

Post by asd-lqw » Mon, Jun 08, 2026, 1:24 PM UTC

Bruce Schneier post is flagged by the Anthropic people.

Limiting the reach of unwanted information is perhaps automated by now. We are further sliding into totalitarianism.

Re: Anthropic's Project Glasswing Update

#7

Post by pseudohadamard » Tue, Jun 09, 2026, 8:58 AM UTC

jerrythegerbil wrote: post #5

Glasswing and Mythos are wildly impressive. The team writing about it has a core charter to publish research about how AI will be disruptive to certain industries. The publication of such research is the disruption. What remains when you stop gamifying the lag time of putting onus of counter evidence of impact and not just minmaxxing the discovery of bugs at the start of a development process is… Does anyone remember…

ITYM "the mythos around Glasswing is wildly impressive". Which is true, look at Anthropic's proposed valuation.