BeyondTrust says its new AI Agent Security tool is available now and checks what an AI helper may do on a work computer before it acts.
BeyondTrust, a security company, announced that its AI Agent Security tool is now available. It is meant to limit what AI helpers can do on work computers before they do it. It runs on Windows and macOS machines through the company's Pathfinder Platform.
The company's concern is simple. AI agents often act for employees with the same access as the person using them, according to its chief technology officer, Marc Maiffret. The tool treats the person and the agent separately. The employee keeps the permissions needed for the job, while the agent gets its own set.
Many security tools block known bad commands or spot risky behavior after it happens. BeyondTrust says it checks each AI agent action against policy while it is running, before the action goes ahead. The company says its research team, Phantom Labs, built the tool with the team behind its Endpoint Privilege Management product.
The launch came with Phantom Labs research on AI coding-agent activity, based on early customer samples. The figures were not in the material available, so they are not reported here. The company notes that its shell command counts reflect commands agents requested, not necessarily ones that succeeded. Pricing and the list of supported AI tools were not stated.
If staff use AI coding or assistant tools, those tools may hold the person's full access, so decide what each agent should be allowed to do.
Visa tested 22 AI models on realistic shopping tasks. They passed just over half, and often bought the wrong thing anyway.
Read more: Visa tested 22 AI shoppers, and most still make costly mistakesDARPA wants AI to hunt for design flaws in stock, bond, energy and cloud markets, so enemies cannot use them to cause economic damage that reaches households.
Read more: Pentagon wants AI to find weak spots in stock and energy marketsTuring Award winner Richard Sutton says more computing power alone will not make AI truly smart. Big AI companies disagree, which matters for anyone betting on AI.
Read more: AI pioneer says piling on computing power won't make machines smart
