Security1 hSeptember 22, 2026, 20:05

AI helps attackers more than defenders, says UK's NCSC

NCSC chief technology officer Dave Chismon said defenders cannot yet use AI as freely as attackers. Attacks have a clear success signal while defensive actions do not, and a wrong automated move can disrupt the very systems being protected. The agency is building its Cyber Shield agentic AI defense program.

The Record Open source

Chismon: attacks have a clear success signal, defense does not
A wrong automated defensive action can disrupt live systems
NCSC is building Cyber Shield agentic AI for government networks
Agentic defense is not ready and will take years of research

#NCSC#GCHQ

Cisco Talos finds CLOSEDQUORUM malware that picks its next move via four LLMs

Security

Elsevier domains briefly redirected to LAPSUS$ "Chapter II" page

Security

Darktrace makes SECURE AI generally available to monitor enterprise AI use

Security

AI helps attackers more than defenders, says UK's NCSC

Cisco Talos finds CLOSEDQUORUM malware that picks its next move via four LLMs

Elsevier domains briefly redirected to LAPSUS$ "Chapter II" page

Darktrace makes SECURE AI generally available to monitor enterprise AI use

Barracuda launches AI Data Security to control employee chatbot use