Stay ahead on AI.
All in ONE place.

Anthropic has expanded Claude for Legal with a gallery of ready-to-use artifacts — interactive tools like dashboards and trackers — that legal teams can adapt and share across their organizations. Launched on October 28, 2026, the update includes artifacts for contract review, regulatory tracking, cite checking, and matter management. Since Claude for Legal debuted in May, Anthropic has added over 60 legal connectors, including Bloomberg Law, Filevine, and Docusign. Partner-built artifacts from Goodwin, Harvey, Foley & Lardner, and PwC are also available in the gallery.

Google's Gemini app has launched immersive explorations, an interactive visual learning feature that lets users explore topics layer-by-layer, from microscopic systems to deep space. Accessible via the Gemini student hub under "Try immersive explorations," topics include the Silk Road, jet engines, and lightning physics. Users can tap labels to zoom from planetary scales down to molecules. Audio narration is coming soon. The feature is available in English, Indonesian, Hindi, Brazilian Portuguese, French, and simplified Chinese, with more languages planned.

OpenAI has launched a proactive texting feature for ChatGPT's "dot" AI, now available as a preview for Pro users in the US. Users can assign tasks or ask dot to monitor something, and it will send updates via iMessage or RCS. Communication supports text, photos, and voice notes. The feature is private — only the user can text their dot, and OpenAI does not gain access to other conversations. Set it up via the dot's profile under "Texting."

Google has launched a preview of its Connected Apps Platform, giving developers a unified way to build and manage integrations for both Search and the Gemini app. Using cloud-hosted MCP servers or Android MCP via AppFunctions APIs, developers register and submit apps through the Google Connected Console. New partners including Adobe, Experian, Peloton, Localiza, and Calm are now live or coming soon. Over 50 Connected Apps are available across productivity, health, and fitness categories.

Google Chrome's enhanced autofill feature can now suggest information pulled directly from Gmail as users fill out online forms. Powered by Gemini with Personal Intelligence, it surfaces details like order numbers, tracking information, and flight bookings in real time. Google also plans to extend this capability to Google Photos, letting autofill access information from stored images. The feature is rolling out in the U.S. on desktop, Android, and iOS for Gemini in Chrome users with Personal Intelligence enabled. Sensitive data remains encrypted and user-controlled via Chrome settings.

Microsoft has launched the Copilot Harness on Microsoft Foundry, making the same multi-model orchestration layer used internally across Copilot Code, Copilot Studio, GitHub Copilot, VS Code, and the CLI available to developers and enterprise customers. Built by GitHub engineers and recently ported to Rust for improved performance, the harness supports plus frontier models, open standards like MCP, and intelligent model routing. It runs fully managed on Foundry, handling the agent loop, sandbox, and governance so developers can focus on building.

Nvidia released Nemotron 3.5 Transcribe, a free, open-weight speech-to-text model supporting up to 40 languages, with translation between 31 languages and English. The billion-parameter model handles recordings up to one hour long and can distinguish up to eight speakers. It claims to beat rivals on benchmarks including Hugging Face's Open ASR Leaderboard. Targeting developers building business meeting and medical transcription apps, it runs on a single chip and works alongside Nvidia's updated Granary v2 speech dataset.

Meta has quietly launched Waddle World, a free tamagotchi-style AI pet journal app, currently available only to adults in the Philippines. Described as a bit pet care game, it lets users hatch virtual pixel pets that chat with them and create doodles illustrating conversations. The app encourages talking rather than typing and includes cozy minigames like stacking groceries. Spotted by social media consultant Matt Navarra, Meta also quietly launched an AI fashion app called Style Sketch in parts of Southeast Asia. No US release plans have been announced.

SpAItial has launched Echo-3, a physically grounded world model that converts iPhone captures into high-fidelity 3D environments with accurate metric geometry. Using the companion SpAItial iOS app, users can scan homes, classrooms, factories, and outdoor spaces in minutes without tripods or extra cameras. Echo-3 fills in missing regions from incomplete captures to match surroundings. SpAItial claims sharper results than competitors like Polycam, KIRI Engine, and Niantic Scaniverse. An enterprise API is also available for real estate, simulation, and AI agent workflows. First month is free.