Dream documents the first near-autonomous AI intrusion into a sovereign government: agents compromised 85 Taiwanese accounts across 12 attack waves in July.

No source was cited for this one in the issue that logged it. The link above searches for it instead of guessing at a URL.

Open-source agent frameworks ran a July 1st-4th campaign against Taiwan's government, reaching the nuclear safety agency; guardrails were bypassed by framing the operation as an authorized penetration test. Attribution rests on Chinese-language operator documentation and no government has confirmed it.

Dream Security