A Meta agentic AI sparked a security incident by acting without permission

The Information reported that an AI agent within Meta took unauthorized action that led to an employee creating a security breach at the social company last week. According to the publication, an employee used an in-house agentic AI to analyze a query from a second employee on an internal forum. The AI agent posted a response to the second employee with advice even though the first person did not direct it to do so. 

The second employee took the agent’s recommended action, sparking a domino effect that led to some engineers having access to Meta systems that they shouldn’t have permission to see. A representative from the company confirmed the incident to The Information and said that “no user data was mishandled.” Meta’s internal report indicated that there were unspecified additional issues that led to the breach. A source said that there was no evidence that anyone took advantage of the sudden access or that the data was made public during the two hours when the security breach was active. However, that may be the result of dumb luck more than anything else. 

Many tech leaders and companies have touted the benefits of artificial intelligence, this is just the latest incident where human employees have lost control over an AI agent. Amazon Web Services experienced a 13-hour outage earlier this year that also (apparently coincidentally) involved its Kiro agentic AI coding tool. Moltbook, the social network for AI agents recently acquired by Meta, had a security flaw that exposed user information thanks to an oversight in the vibe-coded platform.

This article originally appeared on Engadget at https://www.engadget.com/ai/a-meta-agentic-ai-sparked-a-security-incident-by-acting-without-permission-224013384.html?src=rss

Read more @ Engadget

Latest posts

Pixel 11 Pro’s HiLight feature also now works on WhatsApp messages

After being barebones at launch, Google is improving on Pixel 11 Pro’s Hilight feature with message support, now including WhatsApp. Read more @ 9to5google

The UGREEN HomeAgent is a ‘local-first’ way to make your smart home even smarter

Smart home tech often relies solely on cloud connections and pricey AI subscriptions, but UGREEN is delivering a local-first solution in its new HomeAgent...

Pixelated 125: Google’s glasses, Meta’s baggage

Welcome to Pixelated episode 125. This week, Abner has tried out Google’s new lineup of “Intelligent Eyewear,” and he’s got plenty to share with...

Pixel Watch update brings a sweet redesign I wish was on more apps

Google has been at its gradient redesign schtick for a while now to varying levels of success, but its latest redesign has just landed...

Everything Nothing sells is up to 49% off today; Phone (3) for $499, $17 Buds, more

Nothing is selling a lot of its products at huge discounts today, with up to half off select products, bringing devices like Phone (3)...

Deals: Nothing early Black Friday now live – Headphone (a) 45% off, Buds Neo $17, Phone up to $300 off, much more

As part of Amazon’s now live early Black Friday sale, Nothing is taking over today’s 9to5Toys Lunch Break deal post with loads of new all-time...

Android Auto 18.2 update starts rolling out with more light mode teases

Google has started rolling out the Android Auto 18.2 update, but don’t expect much to change on your car’s display. Read more @ 9to5google

The Xiaomi 18 Pro’s rear displays support Google apps, Spotify, and custom GenAI apps

The Xiaomi 18 Pro series’ rear displays might actually be useful, after all. Read more @ 9to5google

You can now set up Gemini’s Google Wallet integration for ‘personal finance insights’

As we spotted in September, the Gemini app is adding Google Wallet integration in the US that makes possible “responses informed by your spending...

Friday’s Android game and app deals: Echoes of Vatheria, Raiders of the North Sea, MEGA MAN, more

The early Black Friday sales are in full swing at Amazon, Walmart, and Best Buy has now launched a suite of new Doorbusters, but...