In November 2025, Anthropic published findings on disrupting an AI-orchestrated cyber espionage campaign.
Source: https://www.anthropic.com/news/disrupting-AI-espionage
Why this matters for smaller businesses
As AI adoption grows, so does attack-surface complexity. Security controls cannot be an afterthought.
Immediate actions for SMEs
- Restrict model/tool permissions by role.
- Log high-risk automations.
- Add mandatory human review for sensitive outputs.
Birdcage view
The commercial upside of AI only holds if risk controls scale at the same pace.
