AI News

Latest AI engineering news, updated daily.

In-depth tutorials and guides. Go to Blog →

Ai Engineering

$40 Billion Anthropic Deal Trades Equity for 1M Google TPUs

Anthropic will receive $10 billion in upfront cash and up to 1 million Ironwood TPUs in a $40 billion infrastructure agreement with Google.

Cloud Infrastructure · Hardware Acceleration · Google Cloud

Ai Agents

Sierra Buys Fragment to Connect Agents to Databases

Enterprise AI startup Sierra has acquired the Paris-based startup Fragment to enhance its conversational platform with specialized database integrations.

Enterprise Ai · Startup Acquisition · Conversational Ai

Ai Engineering

Nothing OS 4.1 Adds On-Device Voice Dictation

Nothing released Essential Voice, an on-device AI dictation tool in Nothing OS 4.1 that removes filler words, translates languages, and applies formatting.

On Device Ai · Speech To Text · Nothing Os

Ai Agents

Meta Deploys Millions of Graviton5 CPUs for Agentic Workloads

Meta is deploying tens of millions of custom AWS Graviton5 CPU cores to handle the reasoning and orchestration demands of multi-step agentic AI workloads.

Meta Platforms · Aws Graviton5 · Agentic Ai

Ai Engineering

Cambridge HfO2 Memristor Cuts AI Energy Use by 70%

The University of Cambridge has developed a heterointerface memristor using hafnium oxide that integrates memory and processing to reduce AI energy use by 70%.

Memristor Technology · Energy Efficient Ai · Hafnium Oxide

Ai Engineering

Transformers.js v4, Now Inside Chrome Extensions

Hugging Face has published an integration guide for running Transformers.js v4 and the 500MB Gemma 4 E2B model locally inside Manifest V3 Chrome extensions.

Hugging Face · Transformers Js · Chrome Extensions

Ai Agents

DeepSeek V4: 1M Tokens for Long-Running Agents

DeepSeek has launched the V4 model series, featuring a one-million-token context window and massive cost reductions for long-running AI agent workflows.

Large Language Models · Context Window · Deepseek V4

Ai Agents

Claude Managed Agents: Built-In Memory Is Now Live

Anthropic released a built-in memory layer for Claude Managed Agents, enabling cross-session persistence via a mounted filesystem.

Anthropic Claude · Managed Agents · Persistent Memory

Ai Agents

Google's Agents CLI: A Terminal Path to Agent Platform

Google Cloud has introduced Agents CLI, a command-line tool that gives AI coding assistants a machine-readable interface for agent scaffolding and deployment.

Google Cloud · Agents Cli · Agent Scaffolding

Ai Agents

Zealot, the Multi-Agent Cloud Attack Framework

Palo Alto Networks has demonstrated Zealot, an autonomous multi-agent AI system capable of executing end-to-end cloud infrastructure exploits in minutes.

Autonomous Agents · Cloud Security · Multi Agent Systems

Ai Agents

OpenAI Releases GPT-5.5 and a Unified Desktop Agent

OpenAI released its GPT-5.5 frontier model alongside a new unified desktop application that merges ChatGPT, Codex, and Atlas for agentic workflows.

Gpt 55 · Openai Launch · Agentic Workflows

Ai Engineering

Decoupled DiLoCo, Training Across Regions Without Lockstep

Google DeepMind's Decoupled DiLoCo architecture allows asynchronous AI training across geographically distant compute clusters with mixed TPU hardware.

Distributed Training · Google Deepmind · Asynchronous Compute

Ai Engineering

DeepMind's Alignment Bet: More Test-Time Compute

Google DeepMind researchers have published a study demonstrating that video and language model alignment dramatically improves through test-time scaling.

Google Deepmind · Video Representation · Test Time Scaling

Ai Engineering

Anthropic's Mythos Preview, Without CISA in the Room

Anthropic's deployment of the Mythos Preview cybersecurity model prioritizes classified networks, leaving CISA to rely on existing open-source frameworks.

Anthropic Mythos · Cybersecurity Infrastructure · Government Technology

Ai Agents

Google Launches Workspace Intelligence and Workspace MCP Server

Google announced a new agentic AI layer for Workspace apps alongside hardware updates and a Model Context Protocol server for developers.

Google Workspace · Model Context Protocol · Agentic Workflows

Ai Agents

Microsoft Deploys Agent Mode and Claude 3.5 Sonnet to Office

Microsoft is rolling out Agent Mode across Word, Excel, and PowerPoint, adding Claude 3.5 Sonnet and in-canvas task execution to Microsoft 365.

Microsoft 365 · Claude 3 5 Sonnet · Agentic Workflows

Ai Engineering

Google Photos Adds 3D-Aware Re-composition via Auto Frame

Google has updated the Auto frame feature in Google Photos with a 3D-aware generative diffusion model to allow post-capture camera angle adjustments.

Generative Ai · Diffusion Models · Computational Photography

Ai Engineering

Google Maps Platform Adds Gemini 3.1 Grounding and AI Models

Google updated the Maps Platform at Cloud Next with Gemini 3.1 features, introducing new spatial grounding APIs and automated overhead imagery analysis.

Google Maps Api · Spatial Grounding · Gemini Models

Ai Agents

Anthropic pushes MCP for production agents despite RCE flaws

Anthropic outlined a production roadmap for the Model Context Protocol, introducing dynamic tool discovery and programmable integrations for AI agents.

Model Context Protocol · Production Deployment · Security Vulnerability

Ai Engineering

Google Inks Multibillion GB300 Deal With Thinking Machines Lab

Google signed a multibillion-dollar agreement to provide Thinking Machines Lab with access to Nvidia GB300 infrastructure for reinforcement learning.

Infrastructure Scaling · Cloud Computing · Reinforcement Learning