AI cybersecurity risks? Humans'll always be #1

AI commands all the headlines nowadays, but the biggest security story of the week is all about human laziness and poor password habits - just like the good old days.

This week on the Kettle, host Brandon Vigliarolo is joined by US editor Avram Piltch (theregister.com/author/avram-piltch) and security editor Jessica Lyons (theregister.com/author/jessica-lyons) to talk about the Klue breach (theregister.com/cyber-crime/2026/06/22/securit…), which was blamed on a "compromised legacy credential" that ought to probably have been deleted a while ago, which allowed cybercriminals to pivot to the SalesForce environments of hundreds of companies. The incident has caused trouble (theregister.com/cyber-crime/2026/06/25/ex-hunt…) for security firm Huntress, who admitted to the breach early on, and the situation over there wasn't caused by AI either.

That said, AI is playing a role in what's being described as "the summer from hell" by one security professional, but while top-tier AI models are spotting troublesome vulnerabilities, the amount of damage they've managed to cause pales in comparison to what one lazy sysadmin can cause by poorly managing passwords.