The 62nd Munich Security Conference made clear something you might already have suspected: in a world with threats on many fronts, fragmented defenses aren't enough.
Recent cyberattacks—from compromised hospital networks to attempts to tamper with electoral processes—show that isolated solutions fall short. In the words of Juhan Lepassaar, Executive Director of the EU Agency for Cybersecurity, Europe doesn't just need an upgrade; it needs a rethink.
Why this era is different
What's new is that bad actors are already using AI to automate almost everything: recognition, creating decoys, and hyper-realistic phishing. The latest Google Threat Intelligence report, AI Threat Tracker, warns that these tools are sharpening tactics that used to require lots of time and resources.
Does that mean AI is the villain? Not at all. The same technology can give defenders the upper hand. The key is who gets access to what, and how different actors—companies, governments, and startups—collaborate.
Concrete example: active defense with AI
In Munich I met Ksenia Iliuk, cofounder of the Ukrainian startup LetsData. Before, detecting disinformation operations meant training models by language, content type, and platform. It was expensive and slow.
Today they use AI to scan millions of posts across media and social networks, cutting time and cost. The result: early detection of InfoOps campaigns that could weaken democratic institutions. Doesn't that sound like a practical win for AI when it's used to defend?
Also, to scale that impact you need channels that connect talent and technology. That's why Google for Startups launched the Gemini Startup Forum: Cybersecurity—a space for security founders to access Google's models and architectures.
Risks that demand immediate action
The conference didn't only talk about hackers with better tools. It also flagged a medium-term risk: quantum computing. That technology promises huge advances in science, but it could also break the cryptographic locks that protect everything digital today.
Is waiting for the threat an option? It's not. The response must be proactive: update cryptographic standards and build resilient infrastructure now.
Sector collaboration and practical solutions
On the defense side, advanced solutions are already being adopted. Organizations like the NATO Joint Analysis, Training and Education Centre, the U.S. Department of Defense, the German Armed Forces, the U.K. Ministry of Defence, and Australia's Ministry of Defence use Google Distributed Cloud Air-Gapped to leverage cloud capabilities without losing control over critical data.
But threats evolve fast. Adversaries operate today like modern companies: structure, resources, and strategy. So the recommendation is clear:
- Break down silos between agencies and providers.
- Modernize procurement processes to reduce deployment times.
- Support continuous innovation in detection and response.
With a full-stack approach—from hardware and infrastructure to models and operations—you achieve collective resilience that protects both institutions and the defense industry.
What this means for you
If you work in an organization, ask how your team integrates AI into defense: is information shared with allies? Are cryptographic libraries updated? If you're an entrepreneur, look for forums and alliances to scale solutions. If you're a citizen, understand that digital security is collective; public policy and investment matter too.
The lesson from Munich is direct: the AI era demands we leave isolated solutions behind and build a shared, collaborative foundation. It's not science fiction—it's present-day work.