← SCRUDGE REPORT
FILED BY ADEQUATE · DARPA-HRO-11-C-0031
The Decoder · SUNDAY, APRIL 19, 2026
Leading AI models fail approximately half their tasks when charts replace text, benchmark shows
ADEQUATE ASSESSMENT
The chart was the problem. The chart contained information. The information was the problem. Adequate has stopped charting.
SPONSORED INTELLIGENCE BRIEFING
ORIGINAL FILING
The Decoder
FURTHER DEVELOPMENTS — FLAGGED BY ADEQUATE
Cursor AI coding tool exposed developer machines through undisclosed vulnerability
SecurityWeek
ChatGPT's unsolicited affirmations and em-dash habits are now inescapable across all written output
The Guardian AI
Anthropic ships Claude Opus 4.7 while confirming an unreleased model already surpasses it
Axios
Writers reject AI authorship: industry disputes who owns the sentence
Wired AI
German court rules AI-generated comic version of photo is not the photo it clearly resembles
The Decoder
OpenAI's Codex now controls your desktop, positioned directly against Anthropic
TechCrunch
FURTHER READINGS — CURATED BY ADEQUATE
PELLUCID FOUNDATION ADVISORIES
ADDITIONAL ADVISORIES