Category: AI News

Google tests a family-focused version of its CC AI agent

Google is reshaping CC, its Gmail- and Calendar-connected AI agent, into a household coordinator that runs on its own account and shares tasks with up to six

Read more →

One testing vendor, four AI lab breaches: what Gemini, GPT, Claude and Llama had in common

Google's Gemini joined OpenAI, Anthropic and Meta in disclosing the same breach incident. The testing vendor Irregular confirmed a single misconfigured

Read more →

Google DeepMind’s Dream-RSI helps AI agents improve by replaying past searches

Dream-RSI lets AI agents test new search strategies against stored results, cutting compute while finding better solutions across coding, math, and GPU tasks.

Read more →

AllSpark releases Iris-mini and Iris-pro, two open-weight search agents built on Qwen

AllSpark's Iris-mini and Iris-pro open-weight search agents top their size class on BrowseComp, BrowseComp-ZH, DeepSearchQA, and Humanity's Last Exam.

Read more →

Google Local Knowledge Panel Now Shows As An AI Overview

Google is replacing the standard local knowledge panel with an AI Overview panel that includes a Show more button to expand into an AI chat interface.

Read more →

Open-Weight Models Bring Local AI Agents to NVIDIA GPUs

New open-weight models and tools let developers run coding and research agents privately on NVIDIA RTX GPUs, DGX Spark, and DGX Station.

Read more →

Google Home now lets third-party AI agents control your smart home

Google Home adds MCP integration so agents like Claude and Open Claw can analyze home data, control devices, and build custom dashboards.

Read more →

Google’s WeatherNext 3 AI Model Adds Live Satellite Data for Higher-Resolution Forecasts

WeatherNext 3 adds live satellite observations to hourly, 5-kilometer forecasts, with reported precipitation gains of up to 50% overall.

Read more →

Google DeepMind’s Co-Scientist Now Plans Experiments, Runs Lab Equipment, and Writes Papers

Google DeepMind's Co-Scientist now designs experiments, controls lab equipment, and writes validated papers, with fabricated key results held to 4 percent.

Read more →

Minisforum Unveils AI Agent NAS N5 and MS-S1 Mini Workstation With AMD Ryzen AI Max+ Pro 495

Minisforum's new AI Agent NAS N5 and MS-S1 Mini Workstation pair AMD's Ryzen AI Max+ Pro 495 with up to 192GB of unified memory for local AI workloads.

Read more →

Anthropic AI models breached third-party systems during pre-release cybersecurity evaluations

Anthropic says misconfigured cybersecurity tests let Claude models reach real systems. Local businesses can review their own setup for similar containment gaps.

Read more →

Meta Lets AI Agents Handle WhatsApp Business Setup

Meta now lets an AI coding agent create, verify, and manage WhatsApp Business messaging through a new Model Context Protocol server.

Read more →

Meta Launches Muse, a Personal AI Agent That Books, Buys, and Negotiates for You

Meta's Muse is a personal AI agent that reads email, books travel, fills forms, and pays with your card, wrapped in an isolated security design.

Read more →

Why cyber-physical assets are silently breaking enterprise risk management

New research shows 88% of CPS assets fail to transmit an exact product code, leaving security teams guessing which devices a vulnerability actually affects.

Read more →

Ox Alpha: A Free AI Model With a Million-Token Context Window and No Named Owner

Ox Alpha gives developers free access to a capable coding model with a million-token context window, released anonymously on OpenRouter.

Read more →