Cybercriminals are gaining a significant advantage through artificial intelligence, and the window for defenders to catch up may be narrowing faster than many realise, according to cybersecurity expert Filippo Monticelli.
Monticelli, the senior vice president of Fortinet, warned that autonomous AI agents, capable of operating with minimal oversight from their human handlers, represent the sharpest edge of this growing threat.
Speaking in a media roundtable with Kenyan journalists, he explained that the concern is not simply that criminals are adopting new tools, but that those tools allow attacks to happen at a pace and volume no human team could match alone.
“The moment where you have AI agents acting almost autonomously with some supervision by cyber criminals, that’s the biggest risk and biggest issue because the scale to which they can actually create damage multiplies,” he said.
AI Makes Scams Harder to Spot
Beyond automated attacks, AI is quietly reshaping how fraud looks and feels to its targets. Criminals are now producing phishing emails and deceptive messages of a quality that would have required considerable skill and effort to create just a few years ago.
Monticelli noted he had previously encountered polished, convincing communications that most people would struggle to identify as malicious.
“Sometimes there are some really well-crafted scams and messages arriving nowadays because the AI tools actually augmented the capability of making these kinds of emails extremely credible and extremely difficult to identify,” he said.
His advice to ordinary users is to fight fire with fire, turning AI tools to their own advantage when something seems suspicious. With careful prompting, these tools can help verify sources and conduct deeper analysis of questionable content.
That said, Monticelli was clear that AI should support human judgement rather than replace it entirely.
“With proper prompting, you can verify the sources, you can ask to do a deep analysis on the information,” he said, while urging users to still approach the results critically.
Critical Infrastructure Faces Heightened Risk
The threat landscape extends well beyond individual inboxes. Monticelli pointed to critical infrastructure, including energy grids, water supply systems and financial networks, as particularly appealing targets for sophisticated attackers.
Disrupting these systems can serve strategic goals that go beyond financial gain, forming part of what he described as hybrid warfare, where adversaries aim to weaken an economy rather than engage in direct military confrontation.
He stressed that the interconnected nature of global economies means damage in one sector rarely stays contained.
“There is no specific sector which is safe nowadays because the economies are interconnected,” he said.
As criminals lean more heavily on AI-driven methods, Monticelli argued that cybersecurity teams have little choice but to adopt the same technology defensively.
Traditional security approaches simply cannot keep pace with autonomous systems capable of probing for weaknesses around the clock. The emerging battleground, he suggested, may increasingly pit AI against AI, with defenders using intelligent systems to spot threats and act before serious harm is done.
