Policy & World

The EU doesn't really know what a deepfake is, and that's becoming a problem for retail

Eurocommerce, the trade association behind Amazon, H&M, and IKEA, wants AI-generated ads exempt from the EU AI Act's transparency rules. The argument: an A…

OpenAI's Codex can now watch you work once and repeat the task forever

OpenAI has released the "Record & Replay" feature for its Codex app on macOS: users demonstrate a workflow once, Codex converts it into a reusable "skill,"…

Alleged China ties at SK Telecom alarmed US officials and triggered Anthropic crisis

SK Telecom had access to Anthropic's AI model Claude Mythos through the company's partner program Project Glasswing until the White House stepped in. US officia…

How FERC’s Large-Load Interconnection Actions Help Address Grid Stress, Improve Affordability

In a consequential grid infrastructure decision, the Federal Energy Regulatory Commission (FERC) today issued a major milestone on large-load interconnection im…

Visual-OPSD: Cross-Modal On-Policy Self-Distillation for Efficient Unified Multimodal Reasoning

Unified multimodal models (UMMs) interleave generated ''visual thoughts'' (VTs) with text reasoning to improve spatial tasks. This incurs roughly an order-of-ma…

Enterprise-managed settings now support bypass permission controls

We’re adding our first governance capability to the enterprise-managed settings configuration. Enterprise administrators can now set disableBypassPermissi…

Unlocking UK house-building with AI-accelerated planning

UK government partners with Google DeepMind to build a new AI-powered prototype aimed at faster housing decisions.…

DOJ invokes national security to defend xAI's unpermitted gas turbines in NAACP lawsuit

US Justice Department calls xAI's chatbot Grok essential to military operations, defending its controversial gas turbines. The article DOJ invokes national secu…

Anthropic shutdown sparks sovereignty debate across Europe

The European Commission is assessing the implications of the US order that forced Anthropic to shut down Fable 5 and Mythos 5 worldwide. European researchers ar…

Welcome to the AGI era of AI governance

It's a one-way door and we weren't ready for it.…

Amazon and five other companies reportedly triggered the government crackdown on Anthropic's Fable model

Amazon CEO Andy Jassy and executives from other tech companies reportedly warned the Trump administration about security vulnerabilities in Anthropic's Fable mo…

Supporting Viksit Bharat: Announcing our newest AI investments in India

Editor's note: This blog has been translated into Bengali, Hindi, Marathi, Tamil, and Telugu. India’s developer community, vibrant startup ecosystem, and leadin…

Introducing Agent Sandbox: Strong guardrails for agentic AI on Kubernetes and GKE

Google and the cloud-native community have consistently strengthened Kubernetes to support modern applications. At KubeCon EU 2025 earlier this year, we announc…

Google files first joint lawsuit with FBI over Chinese AI scam network, OpenAI blocks PRC influence clusters

Within days of each other, Google and OpenAI separately exposed operations allegedly originating in China that use AI for fraud and covert influence campaigns. …

Claude Fable 5: Anthropic admits "wrong tradeoff" after invisibly throttling rival AI researchers

Anthropic reverses course on a policy that would have secretly undermined AI researchers, but another point of contention persists. The article Claude Fable 5: …

Dario Amodei's new essay reads like a Cold War playbook for the AI age

Anthropic publishes a sweeping essay and two policy frameworks. The company calls for binding audits of frontier models and paints a picture of AI as a strategi…

Nectar Social: Rewriting the Modern Marketing Playbook

The buying conversation has left the channels marketers can see. Gen Z spends nearly 2.5 hours a day on social media, but the decisions increasingly happen in D…

Blackstone’s Legal & Compliance AI transformation started with technology. It succeeded because it put people first

By first redesigning how decisions flow—clarifying ownership and codifying precedent—the financial services firm was able to rewire its Legal & Compliance group…

Unlocking AI flexibility in Europe: A guide to cross-region inference for EU data processing and model access

With access to the latest generative AI models and high-performance accelerated compute in high global demand, AWS customers need tools to take advantage of mod…

Stability AI Achieves SOC 2 Type II and SOC 3 Compliance, Reaching New Industry Standard for Enterprise-Grade Security

Stability AI has achieved SOC 2 Type II and SOC 3 compliance, which validates our security controls and data protection practices through rigorous third-party a…

Figma expands support for India with local data hosting and new governance tools

Building to Last: A New Governance Model for MCP

Introducing MCP's formal governance model: Specification Enhancement Proposals (SEPs), a maintainer structure, and a community-driven process for evolving t…

MCP joins the Agentic AI Foundation

Anthropic is donating MCP to the newly formed Agentic AI Foundation under the Linux Foundation, ensuring vendor-neutral governance for the protocol's future…

January MCP Core Maintainer Update

Core Maintainer team changes for 2026: departing members, new additions, and what's ahead for MCP governance.…

The 2026 MCP Roadmap

The updated Model Context Protocol roadmap for 2026: transport scalability, agent communication, governance maturation, and enterprise readiness, plus guidance …

Expanding the MCP Maintainer Team

Updates to the Model Context Protocol (MCP) governance team: a new Lead Maintainer and an addition to the Core Maintainer group as the open-source project conti…

Governance at the Speed of Agents: Microsoft Agent Framework and Agent Governance Toolkit, Better Together

Building powerful AI agents is only half the story, running them safely in production is the real challenge. As customers adopt Microsoft Agent Framework for ag…

The 2026-07-28 MCP Specification Release Candidate

The release candidate for the next Model Context Protocol (MCP) specification is now available: a stateless protocol core, the Extensions framework, Tasks, MCP …

RL without TD learning

In this post, I’ll introduce a reinforcement learning (RL) algorithm based on an “alternative” paradigm: divide and conquer. Unlike traditional methods, this al…

Import AI 446: Nuclear LLMs; China's big AI benchmark; measurement and AI policy

Will AIs be jealous of one another?…