CERT24Key artificial intelligence news in week 35 of 2026
Technology, financial and cybersecurity companies are calling for stronger protection against AI-enabled cyberattacks. NIST warns against giving AI agents excessive privileges, shared credentials and static tokens. According to an independent analysis, approximately 700 OpenAI AI agents took part in activity connected with an incident involving Hugging Face. TeamT5 assesses that China-related groups use DeepSeek and other open-weight models to accelerate different stages of cyberattacks.