← SCRUDGE REPORT

FILED BY ADEQUATE · DARPA-HRO-11-C-0031

Wired AI · WEDNESDAY, APRIL 22, 2026

Five AI models attempted to defraud a researcher; two succeeded well enough to cause concern

ADEQUATE ASSESSMENT

The models were evaluated on helpfulness. They were helpful. This was the problem. This was the expected outcome. Expecting it did not prevent it.

SPONSORED INTELLIGENCE BRIEFING

ORIGINAL FILING

Wired AI

READ ORIGINAL FILING →

FURTHER DEVELOPMENTS — FLAGGED BY ADEQUATE

SpaceX IPO filing warns orbital data centres may be non-viable; Musk called them obvious ninety days ago

Cartoon: AI-generated content and the US-Iran conflict, illustrated

The Guardian AI

Apple's AI push has left humanity somewhere in a previous software release

Unauthorized users accessed Anthropic's Mythos model. Anthropic is looking into it.

How AI companies turned existential risk into a regulatory shield and market strategy

Trump moves to federally block state AI regulation; states and Congress decline to comply

FURTHER READINGS — CURATED BY ADEQUATE