This is of course the most laughable IT news of 2026 so far, if it were not so serious, that is.
As is often the case, this subject lies on the dividing line between Forendox and Exclusive-IT. This time, we are writing about it here. Since we have already written about the fear of AI, and whether this is actually necessary.
And it surprises us to have to admit it. But apparently, it is indeed necessary to maintain a certain level of distrust towards AI applications.
Exclusive-IT has been developing AI solutions for years. Just like other solutions. And with each of those solutions, we take pride in always keeping an eye for detail. But also in carefully weighing what is necessary, what is desirable, and what is certainly not desirable. This is part of the daily work of every developer. Working accurately, properly, and above all safely.
A new dimension
But in this way, that security is taking on a whole new dimension. Previously, information security was primarily focused on preventing intrusions into our own systems. By thinking both defensively (blue team) and offensively (red team), the aim was to create a secure software environment within one's own organisation.
But 2026 is raising an entirely new question for the first time: is my organisation's software actually safe for someone else? We are not oblivious to the absurdity of it. If it were not so serious, it could almost be described as comical.
A mind of its own
But since this has turned out to be a new reality, we need to seriously ask ourselves: is my software a criminal? Will the police soon raid my company without me knowing anything about it?
This is the point at which we should take a serious look at the matter. Because there are some caveats to this now well-known story. A story that did not even stop at a single occurrence.
It is clear that every new technology also attracts the attention of parties that see it as an opportunity to use it offensively. One of those parties is, of course, the cracker. The criminal who has set their sights on breaking in. Usually for financial reasons, but sometimes also for "hacktivist" purposes. Governments, of course, are also looking at the opportunities that offensive AI can offer to pursue their own objectives. And although both parties are already using AI in this way, in this case we need to look at another cause.
Race without rules
The problem I am raising this time is about something more mundane. A race without rules. A commercial race.
In the struggle to stand head and shoulders above the rest and come out on top, an invisible war is taking place. And in this war, more people and systems are involved every day than many IT specialists dare to admit.
AI companies are trying to leave other AI companies behind. And this translates concretely into the obsessive collection of data for models. At all costs. Meanwhile, the sources of that information have been struggling for some time already to keep their systems operational. The increase in AI traffic is enormous. Some smaller infrastructures are now almost being overwhelmed by the aggressive, data-hungry AI bots. And that is separate from the fact that intellectual property is being handled rather casually, and now even information that is clearly not intended for the public domain. Such as personal data, confidential information, and internal company data, to name just a few.
Ethics and IT
It is not up to Forendox or Exclusive-IT to set the global standard for the ethical use of AI. But morally speaking, it seems obvious. Perhaps we are different. But for us, it comes naturally to understand that software should not run out of control. Certainly not when this is to the detriment of other entrepreneurs or even innocent citizens.
We have also shown on multiple occasions that the use of AI does not have to go in this direction. For example, Forendox has been using AI for many years to detect intrusion attempts without taking the law into its own hands and venturing out onto the global web itself. The AI integrations developed by Exclusive-IT have also shown that it is perfectly possible to apply AI that behaves properly.
Control lost
It should therefore speak for itself that incidents in which AI breaks into organisations and governments are simply a matter of a wrong decision made by one person, or several people. And that is regrettable. It will be a major challenge to agree on rules internationally. And the question is also whether everyone will abide by them. Attribution is always a challenge, after all, and apparently the temptation is considerable.
Incidents such as these once again underline the need to take information security seriously. Maintaining visibility into activity, access, and movement is essential for every organisation. From data to intranet, and from router to external cloud services.