Blog

AI in Cybersecurity: Insights from MySecurityEvent Berlin 2026

Written by Matthias Klassen | Tuesday, 12.5.2026

Over the past few days, I had the opportunity to take part in the MySecurityEvent in Berlin and I have rarely experienced a security event that was so in-depth, so practical and so well organized.

Over two and a half days, almost everything revolved around one central question: How is artificial intelligence changing cybersecurity - both on the part of the defenders and the attackers?

AI is massively changing the speed of cybercrime

One particularly impressive point came from the latest 2026 Global Threat Report from CrowdStrike:

The average so-called breakout time is now just 26 minutes - an acceleration of 65% compared to the previous year.

The shortest time measured was just 27 seconds.

The breakout time describes the amount of time an attacker needs to move laterally in the network after initial access and reach critical systems or high-value assets.

These figures show impressively that attackers today work faster, more automatically and more efficiently than ever before - especially through the use of AI and large language models.

 

What attackers are already doing with AI today

Many presentations focused on how aggressively artificial intelligence is now being used by attackers. This was no longer about theoretical future scenarios, but about real developments and methods that have already been observed.

The following topics were discussed, among others:

  • Automated Exploit Discovery
  • AI-generated malware
  • Deepfake social engineering
  • Hyper-personalized phishing
  • Automated reconnaissance
  • Automated infiltration and persistence
  • "Battle of AIs" - i.e. AI against AI in cyber defense

What was particularly exciting was the realization that this is just the beginning.

Many companies are still at the beginning of their AI strategy, while attackers are already experimenting and automating on a massive scale.

 

Security must become faster

Another key point of many sessions was the speed of modern security processes.

Traditional processes simply no longer work in many areas:
Change processes that take days or even weeks, slow response chains and manual checks no longer fit in with a threat situation that changes within minutes.

The clear message from many speakers was therefore

  • scan continuously,
  • continuously assess,
  • continuously adapt,
  • and, above all, automate.

In future, security must think much more in terms of real-time capability.

Or to put it another way: If attackers can act in seconds, defenders must not wait until days later to react.

 

How companies can keep up

In addition to the risks, there was also a lot of focus on how companies can react to this development.

It was emphasized several times that future security strategies will be strongly AI-centric:

  • AI-driven SOCs
  • Automated response processes
  • faster deployments
  • greater use of native OS mechanisms
  • better use of existing telemetry data
  • continuous training based on real SOC results

One sentence in particular stuck with me:

"Stop putting telemetry on a diet."

The message behind it:
Many companies are reducing their database for cost reasons - yet it is precisely this wealth of data that will be crucial in the future in order to be able to use AI-supported defense effectively at all.

Equally exciting was the call for more room for maneuver for security teams:

"Get a mandate to respond and automate ... respond first, ask questions later."

Because when attacks can escalate in seconds, speed becomes the most important defense component.

 

Strong event format with real added value

In addition to the technical topics, I was particularly impressed by the organization of the event.

The event kicked off with a pitch night at which various start-ups presented their solutions - accompanied by investors and potential partners. This was an extremely successful format because you could get to know innovative companies at an early stage and prepare yourself specifically for later discussions.

The two actual days of the trade fair were excellently structured: Presentations, exhibitor areas, networking and so-called "hotseats" alternated continuously.

These hotseat sessions were particularly valuable: you were specifically brought together with different companies and had short, intensive exchange sessions before moving straight on.

This resulted in an incredible number of exciting discussions, new contacts and valuable perspectives in a very short space of time.

 

Cybersecurity in transition

The MySecurityEvent showed very clearly that cybersecurity is currently changing faster than ever before - and artificial intelligence is both an accelerator and a potential protection mechanism.

The coming months will be strongly influenced by how quickly companies can modernize and automate their security processes and integrate AI in a meaningful way.

For me personally, the days at the MySecurityEvent were extremely valuable: lots of new ideas, exciting discussions, strong presentations and, above all, the exchange with people who are intensively involved with the challenges of modern cybersecurity.

We at TestSolutions were very happy to be there - and are already looking forward to the next time.

 

Feel free to get in touch.