All Updates
Page 1735 of 1959
March 5, 2026
Pentagon Labels Anthropic First US Supply Chain Risk
The US Department of Defense has officially labeled Anthropic a supply chain risk, making it the first American company with this designation. Despite the label, the DoD continues to use Anthropic's AI in Iran.
House Advances Meta-Google Kid Safety Bills
A key House committee advanced bills to enhance children's online safety on platforms like Meta and Google. This injects momentum into US regulatory efforts as global countries push for social media protections for minors.
CH Robinson AI Lifts Profits in Disruptions
CH Robinson CEO Dave Bozeman credits AI integration for boosting the firm's bottom line amid Iran conflict supply chain disruptions. The logistics company is praised by investors for handling global challenges effectively. Discussion featured on Bloomberg Tech.
Copilot Hits 60M Code Reviews
GitHub Copilot has reached 60 million code reviews and counting. This feature helps teams manage rapid code changes driven by AI acceleration. The milestone appears on the GitHub Blog.
Claude Tests Memory Import from ChatGPT/Gemini
Claude now supports importing memory—summarized user info—from other chatbots like ChatGPT and Gemini for easy migration. The article details a hands-on test of this new feature. This simplifies switching AI services while retaining conversation context.
Apple Cuts Mac Studio 512GB RAM Option
Global DRAM shortage forces Apple to remove 512GB unified memory upgrade for Mac Studio. Maximum memory now capped at 256GB. Impacts high-end configurations.
Oura Acquires AI Gesture Tech for Ring 5
Oura acquired an AI-driven gesture recognition company. This move could enable voice or gesture controls in the upcoming Oura Ring 5. It hints at advanced AI integration in future wearables.
Coruna Exploit Endangers iPhones
Google’s Threat Intelligence Group and iVerify uncovered the Coruna exploit, compromising iPhones on iOS 13 to 17.2.1 via website visits alone. It installs root-level spyware for data theft, patched in iOS 26, but 26% of recent iPhones remain vulnerable.
Roblox Launches AI Chat Rewriter
Roblox has rolled out a real-time AI chat rewriting feature that automatically replaces inappropriate words with polite, compliant alternatives while preserving user intent. It ditches disruptive '#' censorship for seamless conversations. This maintains community standards without breaking chat flow.
US Plans Global AI Chip Export Controls
US drafting regulations to restrict AI chip exports worldwide without approval. Nvidia and AMD face major impacts. Aims to control global AI training facilities.
Trump secures data center power pledge
Trump has convinced data center companies to pledge funding their own power generation. The pledge lacks enforcement and faces economic doubts, potentially limiting impact.
AI War Risks via Claude Highlighted
Bloomberg columnist Parmy Olson warns that the unclear role of Anthropic’s Claude AI in a US bombing of Iran poses serious problems. Lack of transparency on AI's military use raises ethical concerns. The opinion piece urges scrutiny of AI deployment in warfare.
Broadcom Targets Nvidia AI Dominance
Broadcom issued an ambitious AI sales forecast challenging Nvidia in the chip race. This could reshape AI hardware competition. Bloomberg Open Interest also covers AI portfolio risks and other market news.
Magnificent 7 Vow to Block AI Energy Cost Hikes
Seven top US AI companies and hyperscalers pledged to protect American consumers from electricity price hikes caused by surging datacenter energy demands from the AI boom. The agreement aims to shield households from infrastructure cost increases. Enforcement remains challenging with unclear penalties.
Meta Sued Over AI Glasses Privacy Claims
Meta faces a class action lawsuit alleging false privacy advertising for Ray-Ban smart glasses, after reports of Kenyan contractors reviewing user footage. The suit claims undisclosed human review for AI training misled buyers. Meta confirms contractors filter shared multimodal data for AI improvement.
Luma Launches Creative AI Agents
Luma has introduced Luma Agents, powered by its new 'Unified Intelligence' models. These agents coordinate multiple AI systems to generate end-to-end creative work across text, images, video, and audio.
LangChain Builds Coding Agent Skills
LangChain is developing skills to help coding agents like Codex, Claude Code, and Deep Agents CLI integrate with its LangChain and LangSmith ecosystems. This effort aligns with industry-wide exploration by companies to enhance agent capabilities in coding tasks.
Pentagon Flags Anthropic as Supply-Chain Risk
The Pentagon has formally notified Anthropic PBC that it considers the AI company and its products a risk to the US supply chain. This determination was confirmed by a senior defense official. The move highlights growing scrutiny on AI firms in critical infrastructure.
OpenAI FinTools Rival Anthropic Office AI
OpenAI released a new flagship AI model and financial-services tools optimized for office work. This escalates competition with Anthropic amid their Pentagon risks.
GPT-5.4 Beats Humans 83% on Pro Work
OpenAI's new GPT-5.4 outperforms humans by 83% on professional-level tasks in internal tests. It also shows improved reliability with 18% fewer errors and 33% fewer false claims than GPT-5.2.