AI to play offense, defense in cybersecurity landscape, expert says
Cybersecurity could be transformed into a battle between autonomous AI systems, with artificial intelligence agents conducting cyberattacks and defending organizations with minimal human involvement, a senior Kaspersky researcher has warned.
- Tech
- Anadolu Agency
- Published Date: 11:54 AM 10 October 2026
Cybersecurity is heading toward a fundamental transformation as artificial intelligence agents become increasingly capable of carrying out attacks and defending organizations with almost no human intervention, according to a senior Kaspersky researcher.
"I think there will be a dramatic change. We will essentially see AI fighting AI, with almost no human involvement," Sergey Lozhkin, head of the Global Research and Analysis Team for the Asia-Pacific (APAC) and Middle East, Türkiye and Africa (META) regions at Kaspersky, told Anadolu.
Lozhkin said AI would play an increasingly decisive role in both offensive and defensive cybersecurity operations, with future cyberattacks expected to become largely automated.
"We will see a large number of fully automated attacks covering the entire process, from the initial gathering of information and preparation for an attack to intelligence collection, social engineering, and red-team activities aimed at gaining access," he said.
He explained that AI agents could eventually manage entire cyberattack operations, from initial reconnaissance and intelligence gathering to infiltrating systems, deploying malware and extorting organizations over stolen data.
"The process will then continue with malware infiltration and data extortion. All of this will be carried out entirely by interconnected AI agents operating within a single framework or across several different frameworks," he said.
"These operations will be conducted with an extremely high level of effectiveness and almost no human interaction."
Lozhkin said the rapid evolution of AI technology in recent years suggested that cybersecurity was moving toward a future in which autonomous systems would confront one another.
"Protective mechanisms at the enterprise level will operate in much the same way. Swarms of highly trained and highly capable AI agents will defend organizations against attacking agents," he said.
"It will be like something from a movie -- a battle between AI agents."