> Nuclear-Sabotage Malware Benchmark Trips Up Most Frontier AI Models
[AUTHOR: Eduard Kovacs]
[DATE: 23/07/2026 12:42]
[LANGUAGE: EN]
SentinelOne’s new benchmark, built on the Fast16 case, shows which AI models can sustain a malware investigation and which cannot.
The post Nuclear-Sabotage Malware Benchmark Trips Up Most Frontier AI Models appeared first on SecurityWeek.