軟體韌體更新｜主題 | LAETimes 低空日報

Latest

Researchers Expose 'HalluSquatting' — A New LLM Attack That Tricks AI Coding Assistants into Installing Malware

Researchers from Tel Aviv University and collaborating institutions have disclosed a novel large language model (LLM) attack technique called HalluSquatting. Attackers identify external resource names — such as package repositories or AI agent skills — that LLMs are prone to hallucinating, register those names in advance, and embed adversarial prompt content within them. When an AI coding assistant fetches these fabricated resources, it may execute attacker-controlled malicious instructions.

AI自動駕駛電腦視覺

16 days ago

Source: iThome 台灣

軟體韌體更新

SpaceX Open-Sources Grok Build CLI After Privacy Scandal, Pledges Zero Data Retention

Zoox Issues Software Recall After Robotaxi Sensor Confusion in Smoky Conditions

DJI Releases Avata 360 Firmware v01.00.0500 to Fix FPV Black Screen Bug

Microsoft Patches 570 Security Vulnerabilities in a Single Month, Setting a New Record

Researchers Expose 'HalluSquatting' — A New LLM Attack That Tricks AI Coding Assistants into Installing Malware

Microsoft's Patch Tuesday Set to Expand as AI Accelerates Security Vulnerability Fixes

Xtra Muse Hits All-Time Low of $329, Offering a Budget Alternative to DJI Osmo Pocket 3

DroneShield Q3 Software Update Boosts Counter-UAS Performance Against Faster, Harder-to-Track Threats

Boston Dynamics Launches Spot & Orbit 5.1: New Perspectives Redefine Facility Inspection

Decades-Old Bash Tricks Expose AI Coding Agents to Supply Chain Attacks

DJI Mimo 2.10.10 Update Is Built Around a Camera Americans Can't Buy

Philip Bloom Exposes Insta360 Luna Ultra Auto ISO Image Degradation Bug; DJI Osmo Pocket 4 Unaffected