Agents surge; EU AI deadline nears

Anthropic / Claude ecosystem

No significant new developments.

Frontier model providers

OpenAI Launches Global Partner Network to Scale AI Solutions

OpenAI announced the OpenAI Partner Network, enabling partners worldwide to build, sell, and deliver AI solutions with OpenAI's technology. This network aims to expand enterprise reach and accelerate AI adoption across industries.

Meta Introduces AI Mode on Facebook to Synthesize Answers from Public Posts

Meta launched AI Mode on Facebook, which surfaces synthesized answers from public posts, Groups, and Reels, changing how users search and interact on the platform.

Meta and Prada Partner on AI-Powered Fashion Eyewear

Meta and Prada announced a collaboration on AI-powered fashion eyewear, blending AI with consumer fashion, as Zuckerberg attended the Milan show.

Kimi K2.7 Code: Free Open-Source Model Takes On GPT-5.5

Moonshot AI released Kimi K2.7 Code, a 1T parameter MoE open-source coding model with 256k context, offering near-zero-cost coding assistance.

Amazon Bets on AI-Powered Seller Assistant to Streamline Operations in India

Amazon introduced an AI-powered Seller Assistant in Seller Central for Indian sellers, using Bedrock, Nova, and Claude to automate operations.

Google Demonstrates Long-Running AI Agents That Pause, Resume, and Retain Context

Google showcased durable AI agents on its Gemini Enterprise Platform, capable of pausing and resuming tasks with full context, using three architectural patterns.

AI developer tooling & infrastructure

Kakunin Releases Cryptographic Compliance Shield for AI Agent Ecosystems

Kakunin launched a cryptographic compliance shield for Google Gemini and OpenAI agent ecosystems, providing real-time auditing protections to meet EU AI Act and MiCA standards.

OpenRouter Fusion Launches Multi-LLM API at Half the Cost

OpenRouter Fusion offers a multi-model API that blends responses from various LLMs, reducing costs by roughly 50% through per-model charging and a four-stage processing pipeline.

ValidMind Launches Atryum Open Source Control Layer for AI Agents

ValidMind released Atryum, an open-source control layer for AI agents, and opened early access to ValidMind Agent Authority for enterprise governance of agent autonomy.

Polis Protocol v2.0 Coordinates AI Agents with Markdown-Based Reservations

Polis Protocol v2.0 is a local-first control plane that uses a _polis/ directory and deterministic reservations to prevent collisions and amnesia when multiple AI agents work within a repo.

Kimchi Coding Adds MiniMax M3 Frontier Open-Weight Model

Cast AI's Kimchi Coding autonomous coding agent now offers MiniMax M3, delivering high performance at lower cost.

Cloud & platform providers

AWS DevOps Agent Expands with Custom SRE Agents and MCP/A2A Protocols

AWS DevOps Agent now supports custom SRE agents, headless access, and integrations via MCP and A2A protocols, expanding automation capabilities.

AI policy, regulation & governance

OpenAI Wins Dismissal of Trade Secret Lawsuit by xAI

A US court dismissed with prejudice the trade secret claims brought by xAI against OpenAI, narrowing a key legal dispute between the two AI rivals. The ruling strengthens OpenAI's legal position.

G7 Privacy Regulators Plan AI Enforcement Coordination Before August Deadline

G7 data protection authorities will meet in Paris to coordinate enforcement of AI regulations with 48 days until the EU AI Act deadline, aiming to reduce regulatory arbitrage.

US Senate Passes AI Accountability Act in 68–29 Vote

The Senate passed the AI Accountability Act, mandating documentation, incident reporting, and prohibiting red‑team withholding renewals, sending the bill to the House.

EU Investigates US Control Directive on Anthropic AI Models

EU officials are examining the implications of the US export-control directive that forced a global shutdown of Anthropic's top models, prompting calls for European AI sovereignty.

Industry & market moves

Mark Zuckerberg Admits Meta Made AI Workforce Transformation Errors

Zuckerberg conceded that AI-driven workforce changes were errors and outlined remediation plans while denying imminent broad layoffs.

Cohere Triples UK Footprint with New London Office

Cohere expanded its UK presence with a new London office, aiming to access European AI talent and position itself as a sovereign AI provider for government and regulated sectors.

Jensen Huang Skips Japan, Stirring Anxiety Over AI Leadership in Tokyo

Nvidia CEO Jensen Huang omitted Japan from his Asia trip, while advancing collaborations with South Korea's SK Group on AI factories and semiconductor automation.

Cloudflare Grows AI Team by Hiring Ensemble AI Talent

Cloudflare acquired talent from Ensemble AI to accelerate its AI infrastructure and model efficiency initiatives.

Sarvam AI Raises $234M Led by HCLTech at $1.5B Valuation

Indian AI startup Sarvam AI raised $234 million in a funding round led by HCLTech, signaling strong investor confidence in India's sovereign AI ecosystem.

Salesforce Signs Definitive Agreement to Acquire AI Agent Startup Fin

Salesforce announced plans to acquire Fin, an AI agent platform, expanding its agentic AI capabilities and SMB deployment options.

SailPoint to Acquire Entro to Enhance Agentic AI Security

SailPoint announced intent to acquire Entro to integrate its AI security capabilities into SailPoint's Agentic Fabric for identity governance.

WALT Labs Joins Anthropic's Claude Partner Network for Enterprise AI

WALT Labs partnered with Anthropic to accelerate production deployment of Claude on Google Cloud, offering governance and security controls.

1Password Acquires Apono for Access Governance Across Humans, Machines, and AI Agents

1Password acquired Apono to extend its Unified Access platform, governing humans, machines, and AI agents with dynamic, intent-based controls.

Undo Lands $37M to Provide AI Agents Runtime Context for Debugging

UK-based Undo raised $37M to expand its AI-assisted runtime debugging platform, embedding runtime recordings into AI workflows.

NewCore Launches with $66M to Build AI Agent Identity Security

NewCore emerged with a $66M launch and a platform for first-class AI agent identities, including dedicated lifecycle, permissions, and revocation.

Trust3 AI Joins NVIDIA Inception to Accelerate AI Data Governance

Trust3 AI joined NVIDIA's Inception program to deploy its unified governance layer across multiple platforms for enterprise-scale AI.

AI product & feature launches

Raven Resonance Previews World's First Ambient Computer at AWE 2026

Raven Resonance previewed Raven Prism, an ambient computer with eye and voice control, at AWE 2026. The Linux-based device is positioned as a new computing paradigm ahead of commercial launch.

NVIDIA Expands Local AI Agent Capabilities with RTX Spark PC

NVIDIA introduced the RTX Spark PC with 1 petaflop AI processing and 128GB memory to run agents locally and securely, with Windows and Linux integration.

C1 Launches Autonomous Worker for Governed AI Agent in Enterprise Identity

C1 released C1 Autonomous Worker, an AI agent purpose-built for identity governance, operating under a policy engine with auditable actions.

Draph Unveils Generative Engine Optimization Solution at VivaTech 2026

Draph introduced a next-generation GEO solution for AI-driven e-commerce search optimization in Europe.

Sakana AI Launches Ultra Deep Research Agent for 100-Page Reports

Sakana AI introduced Marlin, an agent that autonomously generates 100+ page research reports after eight hours of reasoning, targeting business intelligence.

Shield AI and Destinus Demonstrate Autonomous Strike and Teaming Capabilities

Shield AI and Destinus completed a full-mission flight exercise showing autonomous collaborative strike and Hivemind piloting for counter-UAS operations.

Research with immediate practical relevance

DRL-Transformer Solves Large-Scale Open-Shop Scheduling Problems

A Transformer policy trained via deep reinforcement learning generalizes from small to 100x100 scheduling instances, outperforming classical heuristics with low parameters.