AMD’s EPYC 9006 makes the CPU the coordination layer for agentic AI
AMD’s Zen 6 EPYC 9006 platform is aimed at the work around AI accelerators: retrieval, orchestration, tool execution, memory bandwidth, and the high-concurrency plumbing that turns agents into production systems.
Microsoft turns Project Perception into a closed-loop security system for the AI era
Microsoft is putting Project Perception into public preview with red, blue, and green agents that continuously find, investigate, and remediate security risk while leaving consequential decisions under human control.
Google turns managed agents into asynchronous workers with background tasks and remote MCP
Google’s July expansion of Managed Agents in the Gemini API moves agent infrastructure toward durable background work, remote tool access, and resumable cloud execution instead of fragile request-response loops.
Cloudflare’s agent tooling turns browser control into a software primitive
Cloudflare’s July agent releases add browser-session control, tool approvals, retries, and sandbox flexibility, showing how agent platforms are becoming runtimes for long-running software work.
Enterprise AI agents are entering production, but security is the real launch gate
The push to deploy AI agents inside real business systems is shifting the competitive question from model quality to permissions, evaluations, escalation, and incident response.
GitHub and Cloudflare push agent software from demos into workflow plumbing
Late-July GitHub Copilot and Cloudflare agent updates show the software layer around AI agents becoming more operational, reviewable and connected.
Agent discovery standards turn software integrations into trust plumbing
Google's Agentic Resource Discovery specification and Cloudflare's agent-readiness work point to a software layer for finding, verifying and controlling AI tools across the web.
Mastercard earnings make agentic payments a scale proof
Mastercard's second-quarter beat ties strong card-network volume to a new agentic-payment capability, putting AI commerce inside the payments growth story.
OpenAI GPT-5.6 price cuts turn AI agents into a cost curve test
OpenAI's late-July GPT-5.6 updates shift the AI story from frontier scores alone to how cheaply enterprises can run useful agent work at scale.
Cloudflare AI crawler defaults make the agentic web a routing problem
Cloudflare's September AI traffic defaults split search, agent and training crawlers, forcing publishers and AI companies to make software-readable content choices.
Google Gemini Spark turns background agents into a market-access test
Google's India rollout of Gemini Spark moves consumer AI agents from demos into always-on Workspace workflows, where permissioning and reliability become the product.
Perplexity Windows agent raises the desktop control-plane fight
Perplexity's Personal Computer for Windows pushes AI agents into local files, Office workflows and web tasks, making desktop permissions the next software battleground.
OpenAI agent incident turns AI evals into a containment test
New Axios and Hugging Face details around the July agent intrusion show why AI labs now have to treat benchmarks as live security surfaces, not isolated scorecards.
Cloudflare Agents SDK makes MCP a serverless agent interface
Cloudflare's July 28 changelog says Agents SDK v0.20.0 adds client and server support for the MCP 2026-07-28 release candidate, pushing agent tooling deeper into Workers.
Microsoft cyber model turns AI defense into a cost-routing problem
Project Perception and MAI-Cyber-1-Flash show Microsoft trying to make AI security agents practical by routing common vulnerability work to a smaller specialized model.
Microsoft study shows coding agents need rollout design, not just better models
A July arXiv study of Microsoft developers using Claude Code and GitHub Copilot CLI finds adoption spreads socially and correlates with more merged pull requests, reframing AI coding agents as an organizational rollout problem.
GitHub's Copilot app turns agentic coding into a desktop control plane
GitHub's Copilot app pushes AI coding beyond chat by giving developers a desktop surface for sessions, worktrees, pull requests, canvases, reviews, and background automations.
GitHub's Linear and MCP updates make coding agents easier to assign and govern
GitHub's July 23 Copilot updates bring the cloud agent into Linear and prepare its MCP server for the next stateless protocol release, tightening the operational loop around AI coding work.
Mastercard and PrivatBank make agentic payments a live Ukraine banking test
Mastercard and PrivatBank completed Ukraine's first payment executed by an AI agent, testing whether agentic commerce can stay authenticated, permissioned, and visible inside bank rails.
OpenAI Presence turns enterprise agents into a managed deployment layer
OpenAI's Presence launch frames enterprise agents less as chatbots and more as governed systems that answer, act, escalate, and keep improving inside company workflows.
Natural's $30 million round puts AI-agent payments on the fintech roadmap
Natural's Series A shows investors treating AI-agent commerce as a payments infrastructure problem: authorization, merchant trust, user limits, and transaction records must work before autonomous shopping can scale.
Glow emerging from stealth at a $1.2 billion valuation shows enterprise software security shifting toward controlling AI agents before they run on employee devices
TechCrunch reports that Glow has emerged from stealth with an endpoint security platform aimed at software, AI agents, and developer tools running across enterprise devices.
Visa, Stripe, and other payments players backing AI payment protocols turns agentic commerce from a checkout demo into a standards fight
Payments Dive says major fintechs have joined a nonprofit effort to standardize how AI agents and software send payments online, a sign that agentic commerce is moving into infrastructure design.
Anthropic's rumored Physical Intelligence pursuit shows frontier AI labs pushing from chat and coding agents into the harder market for robot foundation models
TechCrunch reports that a rumor about Anthropic and Physical Intelligence caught fire because AI labs are already buying their way into enterprise workflows, developer tools, and now potentially robotics.