A Major Open WebUI Flaw Could Have Let Hackers Steal Your AI Chats
Researchers found a critical security hole in Open WebUI’s Direct Connections feature. If exploited, it could lead to full account takeover. The issue is patched in version 0.6.35.
Researchers found a critical security hole in Open WebUI’s Direct Connections feature. If exploited, it could lead to full account takeover. The issue is patched in version 0.6.35.
The first wave of Xbox Game Pass titles for 2026 is live, featuring games like Star Wars Outlaws and Resident Evil Village. This comes amid a chaotic period for Microsoft’s gaming division. The strategy seems focused on giving overlooked games a second chance.
The new James Bond game, 007 First Light, will feature NVIDIA’s latest DLSS 4 technology. But these graphical upgrades are only for the PC version, leaving console players out.
A viral social media post revealed Google had been using Gmail content to train AI models, leading to a class-action lawsuit. The opt-out process is so complex even security experts got it wrong, highlighting major transparency issues.
Nvidia’s CES 2026 announcements are all about software. DLSS 4.5 brings a new AI model to all RTX cards, but the best new feature is locked to the latest RTX 50 series GPUs. GeForce Now is also getting a power boost.
A new OpenAI study reveals a staggering number of people are using ChatGPT for healthcare questions, especially in the U.S. The company sees this as a chance to shape policy, but serious questions about accuracy and safety remain.
A new study from the University of Washington School of Medicine tracked 640 teens and found they spend an average of 1.16 hours per day on their phones during school hours. The research adds fuel to the national debate over banning phones in classrooms.
Microsoft is testing a major visual overhaul of its Edge browser, pulling the design language directly from its Copilot AI. The changes include a settings section that mirrors the Copilot app, rounded corners, and updated context menus. This isn’t tied to a separate “Copilot Mode,” but a fundamental
Experienced software developers thought AI would save them time. In a controlled experiment, it made their tasks take nearly 20% longer. The results challenge the narrative of immediate, widespread productivity gains.
Governments in India, France, and Malaysia are investigating X’s AI chatbot Grok for generating nonconsensual, sexualized synthetic images. India’s IT ministry has given the platform just 72 hours to respond to concerns. The fallout could strip X of its critical legal safe harbor.