The cause for alarm: In recent months, more than 12 states reported possible cyberattacks on water and wastewater utilities—and, as EC contributor and cybersecurity expert Richard Brooks told us, AI makes it “very easy” for hackers to exploit system vulnerabilities within minutes.Â
Digging in: To better understand these risks, cybersecurity company Frenos tested how AI models plan attacks on industrial systems like the software used by water utilities. The “strongest frontier models ” (all unnamed) succeeded around 90% of the time…and this tech continues to advance.
A caveat: The company didn’t test how human intervention could halt or mitigate the attacks.Â
What utilities can do: “Limit the accounts and connections that can reach important systems, watch for changes, and prepare a safe response with engineering and operations,” the report noted. And most importantly: assign a (human) responder to sniff out unusual activity and decide how to act.
Wed, Oct 21
NEWS: Should power pros be worried about AI-fueled cyber threats? Signs point to yes. đź‘€
1