AI for Business

Kaspersky Executive: AI Still Needs a Human Kickoff for Cyberattacks

Artificial intelligence cannot yet launch cyberattacks on its own, according to Anna Kulashova, vice-president for business development in Russia and the CIS at Kaspersky Lab. Speaking at the...

Share:

Artificial intelligence cannot yet launch cyberattacks on its own, according to Anna Kulashova, vice-president for business development in Russia and the CIS at Kaspersky Lab. Speaking at the Tashkent Fintech Forum, she told RIA Novosti that while attacks fully carried out by AI have been observed, an initial human command is still required to set them in motion.

Kulashova pointed to agentic AI as a shift from earlier generative systems, but said the technology has not shown an ability to originate new meaning on its own. She also noted that AI tools can now detect the actions of autonomous agents and track their behavior - for example, by configuring EDR detection and response rules to flag an agent straying beyond its assigned task.

She compared the situation to hiring a contractor who may not be especially bright but is very persistent and takes instructions literally, pursuing the job at any cost unless clear boundaries are set. Once organizations treat AI agents that way, she said, managing them becomes easier. The remarks came on the sidelines of the Tashkent Fintech Forum.

Source: RIA Novosti

Ready to Modernize Your Business?

Get your AI automation roadmap in minutes, not months.

Analyze Your Workflows →