Incident·WS-0055·Aug 25, 2025·disclosed Aug 26, 2025 (1 days later)·in the lab·AI-powered cyberattacks
PromptLock: prototype ransomware run by a local AI model
31/100
WS-2 ModerateHarm score 0, signal score 69. Autonomy: autonomous (the AI ran multi-step operations). How the scores work.
Harm 0/5
damage actually done
Reach 0/5
who was affected
Autonomy 3/4
how much the AI did on its own
Novelty 3/5
how new the behavior was
Permanence 0/5
how hard it is to undo
ESET flagged Go-based samples that use a locally run gpt-oss-20b model to generate Lua scripts on the fly for file enumeration, exfiltration and SPECK encryption, calling it the first AI-powered ransomware. It was later identified as an NYU academic proof of concept ('Ransomware 3.0').
Damage
none realized
Sources
The Hacker News, second source. Correction: academic PoC, not criminal malware in the wild. Uploaded to VirusTotal 25 Aug; ESET disclosure 26-27 Aug; NYU link confirmed ~3 Sept.
Factor ratings and the score are this index’s judgments by a published formula; facts, dates and figures come from the linked sources. Where a lawsuit is pending, the summary reports allegations, not findings.
