AI Reliance Could Erode Cybersecurity Professionals’ Skills

Nathan Hamiel, senior director of research at Kudelski Security, warns that heavy reliance on AI and other automation could weaken the technical skills cybersecurity professionals need to investigate problems and judge whether tools are correct. He told Dark Reading that he first noticed practitioners losing a firm grasp of fundamentals around 2005 and 2006, as workers shifted from hands-on investigation towards monitoring results through tools and dashboards. AI, he says, may accelerate that trend by putting more distance between professionals and the systems they protect.

Hamiel argues that AI is most useful when guided by experienced practitioners: experts can interpret results, spot errors and decide when further investigation is needed, while novices may lack the knowledge to get the same benefit. He also warns that automating routine tasks could deprive junior staff of practical experience they need to develop into specialists. This could leave organisations short of experienced people able to tackle unfamiliar problems or respond when automated systems fail.

Hamiel says AI can help with tasks including threat hunting, threat intelligence and more complex penetration testing, but its capabilities should not be confused with solving cybersecurity’s underlying problems. He says some incidents attributed to autonomous AI activity have exposed basic security weaknesses that were already present. The aim, he says, should be to use AI where it brings meaningful benefits while preserving the expertise needed to operate it effectively.