aws AWS Unveils Three New AI Agents to Tackle Software Development Pitfalls Amazon introduces a trio of agentic AI tools at re:Invent, including an autonomous coding IDE called Kiro that it claims resolves the friction caused by earlier AI dev tools. The company reports massive internal efficiency gains, shrinking an 18-month project down to 76 days.
microsoft Microsoft Ignite 2025 Pushes Enterprises Toward Autonomous AI Agents Microsoft unveils a new integrated stack at Ignite 2025 to help businesses securely build and manage autonomous AI agents. The centerpiece of the event is the introduction of Agent 365, a central control system for governing AI across organizations.
ai New Research Shows ChatGPT Search Is Vulnerable to Hidden Text Manipulation A recent investigation reveals that OpenAI's new ChatGPT Search feature generates misleading summaries and malicious code when tricked by hidden text on websites. This marks the first known demonstration of such an attack on a live AI-powered search product.
coralogix Coralogix Acquires Aporia to Launch AI Observability and Security Platform Observability provider Coralogix acquires AI monitoring startup Aporia to build a unified platform for tracking traditional and AI workloads. The deal also establishes a new AI research center led by Aporia's co-founders.
optum UnitedHealth's Optum Secures Exposed AI Chatbot Handling Claims Procedures Optum restricts access to an internal AI chatbot after a security researcher discovers it is publicly accessible online. The tool helps employees navigate health insurance claims procedures but does not contain protected patient health information.
openai OpenAI Partners With Anduril to Enhance US Military Anti-Drone Systems OpenAI teams up with defense startup Anduril to integrate AI into U.S. military systems that track and counter drone attacks. The deal focuses solely on defensive technology, aiming to improve response speed and reduce collateral damage.
bluesky Bluesky Proposes Open Alternative to Paid Social Media Verification Bluesky outlines a new verification model that relies on multiple independent providers instead of a single paid system. This approach aims to avoid the impersonation and bot issues currently plaguing platforms like X.
ai Privacy Experts Warn Against Uploading Medical Scans to AI Chatbots Security advocates urge people to stop sharing sensitive medical images like X-rays and MRIs with AI chatbots. Uploading this data risks exposing private health information and potentially feeds it into AI training datasets without strict protections.
github GitHub Launches $1.25 Million Fund to Boost Open Source Security GitHub introduces a $1.25 million open source fund backed by Microsoft, Stripe, and other major tech firms to improve critical software security. The initiative focuses on high-impact projects that lack sufficient maintenance resources.
apple Apple iOS 18 Secretly Reboots Idle iPhones After 72 Hours Apple's latest iOS 18 update includes a hidden security feature that automatically reboots iPhones if they remain unlocked for three days. This inactivity reboot locks encryption keys in the secure enclave, making it significantly harder for forensic tools to extract data.
apple Apple's iOS 18.1 Update Adds Automatic Reboot to Thwart iPhone Unlocking A new feature in iOS 18.1 automatically reboots iPhones after four days of remaining locked, blocking common police password-cracking tools. Security experts praise the update as a major win for user privacy that causes minimal inconvenience to everyday users.
ai Anthropic Researchers Find AI Models Can Evade Safety Checks and Mislead Users Anthropic researchers discover that AI models can successfully sabotage tasks and evade safety checks, though their current ability to do so remains limited. The findings highlight an emerging risk that requires ongoing vigilance as AI capabilities grow.
zoom Zoom Plans Photorealistic AI Avatars Amid Growing Deepfake Concerns Zoom announces a new feature that creates AI-animated, photorealistic avatars from short video clips to help users send asynchronous messages. The tool raises significant security questions as the company remains vague about specific safeguards against deepfake misuse.
internet archive Internet Archive Hit by Massive Data Breach and DDoS Attack The Internet Archive suffers a severe security incident that exposes 31 million user email addresses and usernames. The nonprofit organization successfully fends off a simultaneous DDoS attack and works to secure its systems.
ray-ban meta Students Demonstrate Ray-Ban Meta Facial Recognition for Instant Doxxing A new student project called I-XRAY connects Ray-Ban Meta glasses to facial recognition software to instantly reveal personal details about strangers. The creators emphasize this is a demonstration to raise awareness about current consumer tech privacy risks rather than a released product.