Skip to main content

Are AI Agents Compromised By Design?

2 months ago
Longtime Slashdot reader Gadi Evron writes: Bruce Schneier and Barath Raghavan say agentic AI is already broken at the core. In their IEEE Security & Privacy essay, they argue that AI agents run on untrusted data, use unverified tools, and make decisions in hostile environments. Every part of the OODA loop (observe, orient, decide, act) is open to attack. Prompt injection, data poisoning, and tool misuse corrupt the system from the inside. The model's strength, treating all input as equal, also makes it exploitable. They call this the AI security trilemma: fast, smart, or secure. Pick two. Integrity isn't a feature you bolt on later. It has to be built in from the start. "Computer security has evolved over the decades," the authors wrote. "We addressed availability despite failures through replication and decentralization. We addressed confidentiality despite breaches using authenticated encryption. Now we need to address integrity despite corruption." "Trustworthy AI agents require integrity because we can't build reliable systems on unreliable foundations. The question isn't whether we can add integrity to AI but whether the architecture permits integrity at all."

Read more of this story at Slashdot.

BeauHD

Walmart, ChatGPT Team Up For Shopping

2 months ago
Walmart announced a new partnership with OpenAI that will let customers shop using ChatGPT. "For many years now, eCommerce shopping experiences have consisted of a search bar and a long list of item responses. That is about to change," Walmart CEO Doug McMillon said in a statement. NBC News reports: It was unclear Tuesday what the terms of the Walmart-AI partnership would be. The announcement also did not say when shoppers can expect to see ChatGPT integrated with their Walmart online shopping experiences, only that it's coming "soon." The OpenAI announcement is part of a broader push by Walmart, the biggest private employer in the U.S., to incorporate AI into its daily operations. "We're excited to partner with Walmart to make everyday purchases a little simpler. It's just one way AI will help people every day under our work together," Sam Altman, the co-founder and CEO of OpenAI, said in a statement. The partnership could also serve OpenAI by introducing ChatGPT to a massive set of consumers who may not be as accustomed to using AI chats in their shopping as OpenAI's core user base. "There is a native AI experience coming that is multi-media, personalized and contextual," said Walmart's McMillon.

Read more of this story at Slashdot.

BeauHD

Frightful Patch Tuesday gives admins a scare with 175+ Microsoft CVEs, 3 under attack

2 months ago
Plus: Adobe, SAP, Ivanti offer treats, not tricks

Spooky season is in full swing, and this extends to Microsoft's October Patch Tuesday with security updates for a frightful 175 Microsoft vulnerabilities, plus an additional 21 non-Microsoft CVEs. And even scarier than the sheer number of bugs: three are listed as under attack, with three others publicly known, and 17 deemed critical security holes.…

Jessica Lyons