CERT24Key artificial intelligence news in week 36 of 2026Loading data…
CERT24Threat Observatory
Period

Key artificial intelligence news in week 36 of 2026

OpenAI reported that the Astra model reached a critical cyber capability level under its assessment criteria. Anthropic introduced the Claude Fable 5.1 and access-restricted Claude Mythos 5.1 models. According to ESET researchers, the Russia-linked group UAC-0099 used a technique called GuardBreaker to disrupt AI-based malware analysis. In addition, active login sessions of some Claude users were stolen using conventional malware.

← Back to intelligence

Key artificial intelligence news in week 36 of 2026

OpenAI reported that the Astra model reached a critical cyber capability level under its assessment criteria. Anthropic introduced the Claude Fable 5.1 and access-restricted Claude Mythos 5.1 models. According to ESET researchers, the Russia-linked group UAC-0099 used a technique called GuardBreaker to disrupt AI-based malware analysis. In addition, active login sessions of some Claude users were stolen using conventional malware.