Betting that defenders can out-innovate attackers has a mixed track record in security, and nothing about AI so far suggests this round will be different. The recent “outages” hitting the tech and financial sectors aren’t really surprising. But I don’t think AI can take the blame for all of them. AI is more of a catalyst.
Some of the leading AI models have been helping companies identify legacy security flaws and potential exploits at a pace that is a bit too fast for us humans to keep up with. As these holes are uncovered, we need to test and patch them quickly to avoid zero-day exploits. But the sheer volume of these potential problems seems to be overwhelming existing hardware and software, as well as the humans trying to run them.
This “perfect storm” of identification and fixing will continue quickly. So, don’t be surprised if minor and major disruptions continue. The biggest thing is not to spread FUD (Fear, Uncertainty, and Doubt) and to validate and verify as you go.