Huxe, an audio-first app developed by ex-Google NotebookLM team members, offers AI-generated podcasts, daily briefings, and personalized audio content. It recently launched publicly after raising $4.6M.

Google introduced AI-powered photo editing for Android, enabling users to edit images via natural language prompts. Powered by Gemini, the tool supports adjustments like lighting, object removal, and creative enhancements, with follow-up editing capabilities. Initially available in the U.S. for Pixel 10 users, it will expand to Android. C2PA credentials identify AI-edited images.

WhatsApp rolls out in-app message translation, supporting multiple languages on Android and iOS. Users can translate messages via long-press or enable auto-translation for full chat threads. Translations are processed locally to ensure privacy and end-to-end encryption compliance. Meta plans to expand language support and improve accuracy in future updates.

Google has rolled out Gemini AI across its ecosystem, with the Play Store receiving a major update featuring personalized app/game recommendations, the new "You" tab for user activity and subscriptions, AI-generated reviews, localized content hubs, and "Gemini Live" for real-time gaming support. The update also expands Play Games to PCs and improves app discovery, though Google faces ongoing policy-related controversies.

Google expands AI Overviews to Spanish users, offering conversational search, image uploads, and in-depth topic exploration. Rolled out globally in August, it now supports Hindi, Indonesian, Japanese, Korean, and Brazilian Portuguese. AI Ultra subscribers gain premium features like restaurant bookings, with plans for event tickets. Google also expands AI Plus to 40 new countries, competing with OpenAI’s ChatGPT Go, which offers higher usage limits and improved personalization.

OpenAI, with Oracle and SoftBank, plans five new US AI data centers via Stargate, adding to a total planned capacity of seven gigawatts—enough for over five million homes. Three sites, in Texas, New Mexico, and the Midwest, are developed with Oracle, while SoftBank supports two in Ohio and Texas. This expansion supports OpenAI’s growth as it trains more powerful AI models, backed by a recent $100B Nvidia investment for AI chips and data centers.

Meta introduces AI assistant on Facebook dating to enhance match suggestions and profile optimization, amid rising AI integration in dating apps like Tinder and Hinge, which are also investing heavily in AI-driven features.

Oracle reshuffles executive team, promoting Clay Magouyrk and Mike Sicilia to co-CEOs as it strengthens AI infrastructure focus. Safra Catz transitions to executive vice chairman. Recent deals with OpenAI and Meta highlight Oracle's growing AI ambitions.

NVIDIA plans to invest up to $100B in OpenAI, deploying 10 gigawatts of AI hardware to support advanced AI model development, marking the largest AI infrastructure commitment to date.

Gemini AI is now available on Google TV, enabling natural conversations and expanding to 300M+ devices. It helps users find shows, recall plot details, locate movies, and answer general questions. Available on TCL QM9K series and more models in 2025.

xAI launched Grok 4 Fast, a high-performance, low-cost AI model with 2M token context window, rivaling Gemini 2.5 Pro at 25x lower cost. It excels in speed, token efficiency, and reasoning, offering developers affordable, real-time AI capabilities.

Sumo Logic launches Dojo AI, an agent-driven security service automating routine tasks, enabling proactive threat response. Featuring Mobot, Query Agent, and Summary Agent, it enhances productivity by streamlining investigations and improving collaboration.

Volcano Engine unveils Doubao Translation Model, supporting 28 languages with high accuracy and fluency. Matches Deepseek-R1 in Chinese-English translation and rivals GPT-4o and Gemini-2.5-Pro in multilingual performance. Handles complex contexts like poetry, tech, and medical terms, with 4K context window and 3K output length. Priced at 1.20 yuan/million input chars and 3.60 yuan for output.

xAI, Elon Musk's AI firm, raised $10B in new funding, boosting its valuation to ~$200B. The funds will expand tech capabilities, build data centers, and develop the Colossus supercomputer with up to 1M GPUs. The company also plans to grow its team and achieve profitability by 2027.

Oracle's stock rose 4% amid reports of a potential $20B cloud deal with Meta for AI infrastructure, while its partnership with OpenAI drives strong YTD gains and increased capex.

World Labs, led by Fei-Fei Li, unveils Marble, a generative model that creates persistent 3D environments from a single image and text. It builds on "large world models" to go beyond 2D and language AI, generating navigable, exportable 3D spaces using Gaussian splat rendering. While promising for game development, VR, and AI training, it struggles with non-photorealistic or highly imaginative inputs. Marble contrasts with DeepMind’s Genie 3, which simulates interactive worlds without producing s

Luma AI launched Ray3, a text-to-video AI model with reasoning capabilities for cinematic production, doubling its size from Ray2 and supporting HDR video in 10/12/16-bit formats. It features a Chain-of-Thought Reasoning engine, enabling intelligent scene planning, multi-step creative processes, and annotation interpretation. Integrated into Adobe Firefly and adopted by Dentsu Digital, Humain, Monks, and StrawberryFrog, Ray3 enhances creative workflows with improved consistency, HDR conversion,

Google rolls out Gemini features to all Chrome desktop users in the U.S., previously limited to subscribers. The update includes multi-tab support, browsing history retrieval, deeper integrations with Google apps, and AI-powered search in the address bar. Gemini can now assist with trip planning, shopping comparisons, and content clarification. Google also introduces scam detection using Gemini Nano and one-click password resets for compromised accounts. AI Mode enables advanced searches with fo

OpenMind releases OM1, an open-source AI operating system for robots, enabling developers to build, test, and deploy intelligent robotic applications across drones, quadrupedal, and humanoid robots using modular AI models and agent workflows.

Microsoft introduces AI agents for Teams meetings, channels, and communities, enabling real-time task management, decision tracking, and automated document creation. The Meeting Host Agent captures action items, creates tasks in Planner, and drafts proposals during meetings. Channel Agents provide status updates and generate reports, while the Viva Engage Community Agent drafts verified responses. GitHub integration allows assigning coding tasks to AI. Available now to Microsoft 365 Copilot user