We learned to trust AI faster than we learned to defend it

Quint Ketting (Atos) at Cybersec Netherlands 2026

It does not begin with a hacker in a dark hoodie. It begins with a decision that seems logical. A process is automated. An AI agent is given access to company data. An employee allows an algorithm to handle emails or write code independently. Everything becomes faster, more efficient and smarter. Until the moment no one is still asking who is in control when things go wrong.

That is exactly the pattern we have seen before. In the late 1990s, we connected the world through the internet, while security was mostly an afterthought. Only after cybercrime, ransomware and large-scale data breaches exposed the downside of this digital acceleration did security become a discipline in its own right. Now organisations are once again facing a similar tipping point. This time, it is not the internet, but autonomous AI that is rapidly changing the way we work, make decisions and control systems.

Faster attacks with AI

This development is at the heart of the keynote that Quint Ketting, Executive Advisor Cybersecurity at Atos, will deliver on the first day of Cybersec Netherlands 2026, taking place on 9 and 10 September at Jaarbeurs Utrecht. His central question is as simple as it is uncomfortable: are the rules on which cybersecurity has relied for years still sufficient in a world where attacks evolve with the help of AI faster than humans can respond?

That question points to a fundamental shift. For years, cybersecurity has revolved around control. Processes were established, governance was tightened and compliance requirements expanded. Frameworks and certifications created structure and predictability. All of that remains necessary. But when attackers use systems that continuously learn, make decisions automatically and operate without interruption, a new problem emerges: speed.

Carefully designed processes

An attacker no longer has to wait for office hours or a human decision. AI can analyse vulnerabilities, personalise phishing campaigns, adapt malware and explore new attack paths without pause. Traditional defence organisations, on the other hand, often still operate at a very different pace. Alerts are escalated, forms are completed, meetings are scheduled and decision-making follows fixed procedures. These processes are carefully designed, but they are not always designed for an opponent that adapts in seconds.

That does not mean governance loses its value. On the contrary. Good governance, clear responsibilities and demonstrable compliance remain the foundation of digital resilience. The challenge is that governance alone is no longer enough. Organisations will need to find ways to combine speed and control.

A different way of working together

This requires a different way of collaborating. Not only between security teams and IT departments, but also between executives, developers, suppliers and users. Cyber resilience is becoming less a collection of separate measures and more a continuous process in which information, insights and decisions must be shared faster. Periodic checks are giving way to continuous monitoring. Static security measures are being supplemented by systems that can adapt during an attack. And trust is shifting from hierarchical decision-making to teams that are empowered to act immediately when the situation demands it.

That may be the greatest challenge of the AI era. Technology is developing faster than organisations are adapting their ways of working. New AI applications are implemented because the benefits are clear, while the governance, security and collaboration around them often follow later. This risks repeating the same mistake made during the rise of the internet: first innovate at scale, then discover which risks have been created.

Making decisions faster

This calls for a different reflex. Not more rules, but better collaboration. Not less governance, but governance that enables speed. Not only investing in new technology, but also in the ability of people and organisations to make decisions together more quickly when circumstances require it.

The rise of autonomous AI therefore makes cybersecurity not only a technical issue, but also an organisational one. How do you organise trust when systems act independently? How do you maintain control without slowing down innovation? And how do you ensure that security does not become an afterthought in digital transformation, but is part of the architecture from the start?

The next phase of cybersecurity

These are precisely the questions that will define the digital resilience of organisations in the years ahead. During Cybersec Netherlands 2026, Quint Ketting will take the audience through this development on the first day of the event. Not with a call for more fear or more regulation, but with a critical look at how organisations are preparing for an era in which speed, collaboration and trust will become just as decisive as technology itself. Anyone who wants to prepare for the next phase of cybersecurity will need to invest not only in better systems, but above all in a different way of thinking and working together.