This Week in AI

OWASP’s first ranking of the top 10 risks for agentic AI highlights threats such as goal hijacking, unsafe autonomy, and tool misuse.

Telco providers are introducing domain-specific LLMs to automate operations and monetize intelligence across complex networks.

Microsoft and Dell report that enterprises are shifting from experimentation to disciplined execution with AI.

Palo Alto Networks reports that 99 percent of organizations experienced attacks on AI applications last year, with API attacks rising 41 percent.

NIST’s preliminary Cyber AI Profile maps CSF 2.0 to three areas: securing AI systems, enabling AI-powered cyber defense, and countering AI-enhanced attacks.

Solaris AI launches as a four-week accelerator designed to compress typical one-year enterprise AI transformation timelines.