The fusion of AI and cyber is no longer a theoretical risk; it is fast becoming a strategic reality. During the fourth day of AI Week by Commit and Calcalist, in collaboration with Google Cloud, Yossi Karadi, Head of the Israel National Cyber Directorate, warned of a future where a full-scale cyber war could unfold without a single shot being fired.

Karadi outlined a scenario in which a digitally powerful adversary uses AI-driven cyberattacks to impose a complete digital siege on a country. Critical infrastructure, financial systems, health services, and communications could be disrupted in parallel, with ordinary citizens feeling the impact first. In such a world, forcing a state to raise a white flag becomes a matter of sustained digital pressure, not traditional military force.

He described three layers where AI and cyber are already colliding. First, AI is enhancing cyber defense with faster detection, smarter clustering of incidents, and automated responses. Second, AI systems themselves must be secured, including both models and the data that trains them, since poisoned data or manipulated models can quietly distort decisions. Third, we are rapidly approaching a battlefield where AI agents attack and AI agents defend in real time.

While attackers are already using AI to write better code, refine malware, and continuously scan for vulnerabilities, defenders are racing to turn AI into a force multiplier. In Israel, AI is being used to correlate massive volumes of alerts into coherent campaigns and to deploy agents that reduce the pressure on human cyber teams while improving response speed and accuracy.

The strategic message is clear. Nations and organizations that treat AI security as part of their core resilience strategy, not as a niche technical concern, will be better positioned for the cyber conflicts of the near future.

‍