AI WAEPONIZATION

How threat actors are advancing

Agentic AI attacks

Anthropic’s Claude Code tool was hijacked by state-sponsored actors to run autonomous intrusions across 30 government and financial entities, with 80–90% of the attack lifecycle executed without human intervention.

Deepfake social engineering

Real-time video and voice cloning now bypasses biometric verification. The Arup case proved it: “sensory trust” – the assumption that seeing and hearing a colleague means they are real – is dead.

Supply chain gateway doctrine

5 of the 8 top incidents entered via compromised third-party vendors. Hardening your perimeter is irrelevant when your HVAC contractor or HR software provider is the entry point.

Just-in-Time (JIT) polymorphic malware

New malware families (PROMPTFLUX) use LLMs to rewrite their own code in real time, defeating traditional antivirus and EDR systems that rely on recognising known file hashes or code strings.

KEY TOPICS
Breach Cost by Market 2025

WHAT’S INSIDE 
60 Pages of Intelligence for Leaders 

• Forensic analysis of 9 landmark incidents, from the Bybit Heist to the Dodd Group breach
• EMEA breach cost statistics with MTTD/MTTR per market: GR, CY, SA, UK, Benelux, Africa
• Sector Risk Matrix across Banking, Energy, Maritime, Healthcare, Telecom and Government
• Strategic mandate for C-Level executives and government officials, with actionable checklist
• AI weaponization deep-dive: Agentic AI, JIT malware, deepfake fraud, quantum computing threats
• The missing dimensions: cyber insurance exclusions, workforce crisis and ransom economics
• Post-2025 Executive Checklist: Immediate – Structural – Strategic actions
• Appendix: 28 EMEA regulatory and compliance frameworks – NIS2, DORA, EU AI Act and more