Sugiura Takayuki (杉浦隆幸) of the Japan Hacker Association says AI attacks are expected to become 100 times faster than attacks carried out by humans. He explains why patching should be automated, along with priorities and practical steps for responding.
As large-scale data breaches continue to occur, Sugiura Takayuki (杉浦隆幸) believes AI agents are part of the reason, allowing attacks to proceed far more quickly than those carried out by humans. Attackers execute the entire process of reconnaissance, vulnerability discovery, intrusion and data theft in a single operation, making it increasingly difficult for conventional measures alone to keep pace.
Drawing on his own research experience, Sugiura Takayuki (杉浦隆幸) explains that when he incorporated his expertise into an agent for research in March this year, work that had previously taken a month was completed in a few hours. He estimates that AI operates at approximately 100 times the speed of humans because it runs without rest, 24 hours a day, 365 days a year.
A common feature of companies that suffer damage despite having adequate budgets and personnel is their adherence to past practices. Although the situation calls for the introduction of either top-tier talent or AI-powered diagnostics, rather than moderate-level vulnerability assessments, many companies have yet to establish the necessary framework for AI-powered diagnostics. He also points out that relying on outsourcing providers or outsourcing development carries risks.
One point readers should consider in the context of their own companies is whether they can explain the contents of their systems themselves. The key to effective measures, the article concludes, is advanced AI-assisted diagnostics and the establishment of a framework capable of implementing them.