Anthropic is rolling out a new Claude feature that generates charts, diagrams, and visual step-by-step guides as part of conversational responses. The capability uses HTML and SVG code rather than image generation, and is available to all users in beta form on desktop.
Anthropic has made interactive chart generation available to Claude users as a beta feature. The capability, previously shown as 'Imagine with Claude' last fall, allows Claude to generate dynamic visualizations using Chart.js, HTML, and CSS that users can modify in real-time as conversations develop.
GitHub has removed premium AI models including GPT-5.4, Claude Opus, and Claude Sonnet from its free Copilot Student plan effective March 12, 2026. The change leaves students with access only to lower-cost models: Claude 4.5 Haiku, Gemini 3.1 Pro, and GPT-5.3 Codex. GitHub's decision triggered 2,874 downvotes versus 21 upvotes on the announcement, with students arguing they need premium models to learn industry-standard tools.
Microsoft has launched Copilot Health, a separate space within Copilot designed to aggregate electronic health records and wearable data for personalized health insights. The service explicitly disclaims providing medical advice, diagnosis, or treatment—positioning itself as a wellness tool rather than clinical decision support.
Google has restructured pricing for its Antigravity agentic AI coding tool, shifting from quota-based access to a hybrid credits model. Developers on the $20/month AI Pro plan report weekly rate limits have tightened dramatically—one user tracked a drop from 300M input tokens weekly to under 9M. The company now directs heavy users toward the $249.99/month Ultra plan.
Meta has acquired Moltbook, a social network designed exclusively for AI agents, and hired its founders Matt Schlicht and Ben Parr to work in Meta's Superintelligence Labs run by former Scale AI CEO Alexandr Wang. The acquisition gives Meta access to Moltbook's technology for verifying agent identities and coordinating complex tasks between AI bots. The move signals Meta's intent to integrate agentic AI capabilities into its platforms, though specific plans remain undisclosed.
Perplexity announced Personal Computer today, a cloud-based AI agent that runs on a continuously operating Mac mini to merge local applications with Perplexity's AI capabilities. The system operates 24/7, accessible from any device, and maintains integration with users' files and applications through a secure local connection.
Google has expanded Gemini in Chrome beyond the US to Canada, India, and New Zealand, adding support for over 50 languages. The update brings Gemini to iOS browsers and includes new image editing capabilities via Nano Banana 2, while maintaining existing integrations with Gmail, Calendar, Docs, Drive, Maps, Search, and YouTube.
OpenAI has added support for signing into multiple ChatGPT accounts within separate profiles in its Atlas browser. The feature addresses what product lead Adam Fry called "one of the biggest blockers" to using Atlas as a primary browser.
Meta has acquired Moltbook, a social network designed for AI agents, bringing founders Matt Schlicht and Ben Parr into Meta Superintelligence Labs under former Scale AI CEO Alexandr Wang. The move positions Meta alongside OpenAI's OpenClaw in acquiring AI agent platforms.
Nvidia will invest $26 billion over the next five years to build open-weight AI models, according to a 2025 financial filing confirmed by executives. The move signals a strategic shift from chipmaker to AI frontier lab, with the company releasing Nemotron 3 Super (128B parameters) and claiming it outperforms GPT-OSS on multiple benchmarks.
Amazon announced a new "Sassy" personality for Alexa+ on Thursday, marketed toward adult users and protected by additional security checks including Face ID on iOS. The personality uses explicit language and wit but explicitly excludes sexual content, hate speech, and harmful material.
OpenAI has added interactive visual explanations to ChatGPT, enabling students to explore mathematical formulas, variables, and scientific concepts in real time. The feature allows users to manipulate and test variables dynamically rather than relying on static text descriptions.
Descript has integrated OpenAI models to enable multilingual video dubbing at scale, optimizing translations for both semantic accuracy and speech timing to produce natural-sounding dubbed content. The system balances meaning preservation with practical constraints of dubbed audio synchronization.
OpenAI has released Codex Security as a research preview, an AI application security agent designed to detect and patch complex code vulnerabilities. The tool analyzes project context to reduce noise and increase confidence in vulnerability detection.
Google's task automation feature for Gemini is now live in beta on Samsung's Galaxy S26 Ultra, enabling the AI assistant to autonomously complete actions in food delivery and rideshare apps. The system can order rides, food items, and make decisions like warming pastries—stopping before final confirmation for user review.
Facebook Marketplace is rolling out Meta AI features that automatically draft replies to buyer inquiries and generate item listings from photos. Sellers can also now offer shipping on listings and view buyer profile summaries with activity history.
xAI's Grok 4.20 scores 48 on Artificial Analysis' Intelligence Index—6 points ahead of Grok 4 but trailing Gemini 3.1 Pro Preview and GPT-5.4 at 57. The model distinguishes itself with a 78% non-hallucination rate on the AA Omniscience test, the highest recorded across any model tested.
OpenAI released version 2.27.0 of its Python SDK on March 13, 2026, adding significant improvements to the Sora video generation API. The update introduces character API support, video extension and editing capabilities, and higher resolution export options.
AI2 has released MolmoSpaces and MolmoBot, robotics models trained exclusively in simulation that transfer directly to real robots without manual real-world data collection or fine-tuning. The approach eliminates months of teleoperated demonstrations typically required for simulation-trained robots. Both systems are open-source.