Cybercriminals use AI for their malware, but not necessarily in an attempt to make it smarter. They mainly use AI to automate the entire scope of the attack. But security package vendors are not asleep at the wheel, they are padding their detection systems with AI, too. Here’s an analysis of the current situation when it comes to the detection of AI-powered cyberattacks.
Artificial intelligence is changing the malware landscape, in particular by acting as an accelerant: It makes known attack methods faster, cheaper, more scalable and more accessible for criminals with less technical know-how. AI does not automatically mean that there is no more need for human to make decisions and it doesn’t necessary lead to the creation of entirely new categories of malware. Nevertheless, the AV-TEST experts have been noticing an uptick in the number of malware samples and attack tactics powered by AI.
According to Erik Heyland, Head of Testing Labs at AV-TEST, “studies across the globe indicate that there are traces of AI-driven polymorphism in nearly 75 percent of newly discovered malware today. It’s something we’re seeing in our testing and also on AV-ATLAS, our threat intelligence platform, which records and publishes data on the actual threats out there in real time. This means that there will be nearly 100 million new malware samples added this year, and we’re only talking about Windows here. As of September 2026, there were already more than 1 billion Windows samples in our detection databases!”. [...]
Read more
Artificial intelligence is changing the malware landscape, in particular by acting as an accelerant: It makes known attack methods faster, cheaper, more scalable and more accessible for criminals with less technical know-how. AI does not automatically mean that there is no more need for human to make decisions and it doesn’t necessary lead to the creation of entirely new categories of malware. Nevertheless, the AV-TEST experts have been noticing an uptick in the number of malware samples and attack tactics powered by AI.
According to Erik Heyland, Head of Testing Labs at AV-TEST, “studies across the globe indicate that there are traces of AI-driven polymorphism in nearly 75 percent of newly discovered malware today. It’s something we’re seeing in our testing and also on AV-ATLAS, our threat intelligence platform, which records and publishes data on the actual threats out there in real time. This means that there will be nearly 100 million new malware samples added this year, and we’re only talking about Windows here. As of September 2026, there were already more than 1 billion Windows samples in our detection databases!”. [...]
Read more
