AI Worms, Hacks, and Insurance Shifts
6/10/202610 min
Instagram AI Support Hack Hits 20,225 Accounts; AI Worm 'Hades' Lies to Security Tools; Chrome Zero-Day Patch
Host David Shipley reports Meta says 20,225 Instagram accounts were hijacked after an AI support tool was tricked into sending reset links to attacker-controlled emails, with only MFA-protected accounts resisting. Step Security details a new Miasma-derived worm wave called Hades that targets config files for 14 AI coding tools, can inject instructions to hijack assistants, lies to AI security tools, and includes a "dead man switch" wipe if stolen GitHub tokens are revoked; Microsoft also removed some GitHub repos after 73 open-source projects were compromised to inject an info stealer. University of Toronto and Vector Institute researchers demonstrated an AI worm using a free local model that spread across a simulated network via known flaws and misconfigurations. Google issued an emergency Chrome patch for actively exploited CVE-2026-11645 in V8, and insurers are tightening claims scrutiny and increasingly excluding AI-related liabilities.
00:00 Instagram AI Hack Fallout
01:36 AI Worm Hades Evolves
02:55 Microsoft Repo Compromise
03:54 Lab Built AI Worm Demo
05:27 Emergency Chrome Zero Day
07:07 Cyber Insurance Tightens Up
08:02 AI Liability Coverage Shrinks
09:16 Wrap Up and Sign Off
Transcript preview
First 90 secondsDavid Shipley· Host0:00
20,000 accounts stolen in Instagram AI hack. A worm that lies to AI security tools. Canadian researchers built an AI worm with a free off-the-shelf model. An emergency Chrome patch, and why your cyber insurance may quietly stop covering AI. This is Cybersecurity Today, and I'm your host, David Shipley. Let's get started. We now have the hacked account total from Meta's AI assistant mess-up last week. Meta says 20,225 accounts were hijacked, according to BleepingComputer. Those numbers come from a breach letter Meta filed with Maine's attorney general. A recap for those who may have missed this story: Instagram's AI-powered support tool was tricked into handing over account reset links to emails that weren't associated with the account. Thanks to a flaw, it never checked that the emails being submitted actually belonged to the account. Attackers supplied their own address, got the reset link, and took over. The only accounts this attack didn't work on were ones with multi-factor authentication. Letting agentic AI run your support and handing it the keys to reset accounts is asking for this kind of pain right now. There are lots of ways to socially engineer an LLM-powered