CISA Advisories

🚨 Spies Exploit BeyondTrust 0-Day & Meta Muse Spills Your Secrets 🤖

Today's two biggest security stories hit both sides of the fence: a remote-access tool trusted by IT teams was quietly hacked for months, and an AI chatbot's "share" button turned private conversations into public breadcrumbs. Here's what you need to know.

🔓 The Back Door IT Teams Didn't Know Was Open

BeyondTrust Remote Support, software IT departments use to remotely fix employees' computers, has a critical flaw (rated 9.3 out of 10) that lets attackers walk right in without a password. Once inside, they can take full control of the system and reach the vaults where companies store their most sensitive login credentials. A suspected Chinese state-sponsored espionage group has been exploiting it since late August, planting hidden tools so they can stay inside and move deeper into networks. Patches are now out, and U.S. and Canadian cyber agencies are urging organizations to lock down external access and check whether they've already been breached. If your company or IT provider uses BeyondTrust, ask today whether it's been patched, because "fixed now" doesn't mean "never compromised."

Read more

🤖 Your AI Chat Wasn't as Private as You Thought

Researchers found about 6,800 shared conversations from Meta's AI assistant, Muse, that could be discovered online, because each share link put the conversation's title right in the web address. That made them findable with publicly available internet records. The exposed chats included executive hiring notes, revenue reviews, loan applications, bank-fraud details, property records, and staff calendars. Meta paused the sharing feature on October 20 while it investigates. The lesson: treat anything you paste into an AI tool, and especially anything you "share," as something that could one day be public.

Read more

Stay ahead of threats with GOCO Security at gocosecurity.com.

Comments