Read this because This framework solves the limitation of internal reasoning alone by grounding models in external evidence, offering a significant performance boost with lower token costs.
OpenVLHarness improves multimodal model scores by up to 12.6 points using persistent memory and heterogeneous tool integration.
Read this because AI agents often pass tool-call checks but fail to update databases correctly. This benchmark exposes the gap between successful API calls and actual data consistency.
Microsoft and Hugging Face release ThinkingBox, a benchmark grading AI agents on terminal backend state and side effects across 507 workflows.
Read this because Holo4 offers a unified model for multi-interface automation, competing with frontier models at lower cost while open-sourcing training trajectories.
Hcompany introduces Holo4, a new series of agentic models supporting GUI, code, MCP, and API interfaces for generalist computer-use tasks.
Read this because Builders need a common layer to bridge disparate agent stacks, models, and tools without rewriting integration logic.
Ecma International has standardized NLIP to enable interoperability across diverse AI agent development frameworks and execution environments.
Read this because AI engineers building desktop agents can use TMI to learn real-world workflows from raw interaction data, enabling more robust and auditable task execution.
New TMI method extracts structured task models from passive computer-use traces, significantly improving agent accuracy over existing baselines.
Read this because The race in agentic coding has shifted from model quality to runtime. OpenAI is buying the boring, hard infrastructure — persistence, sandboxing, audit — that turns a clever agent into a deployable one.
OpenAI is acquiring Ona — the cloud dev-environment firm formerly known as Gitpod — to let Codex agents run longer tasks in persistent, audited environments.
Read this because The most telling data point: one customer burned its entire 2026 AI budget in four months. Ramp is building the observability and governance layer for AI spend — a category that barely existed a year ago.
Ramp raised $750M at a $44B valuation, betting AI agents will own corporate spend and launching the first corporate card purpose-built for AI agents.
Read this because We skipped the obvious MAI-models headline because the durable Build 2026 story is structural: Microsoft wants grounded context to be a default platform layer, not a feature. Watch Web IQ — a low-token web-grounding broker.
At Build 2026 Microsoft folded grounding into one brand, Microsoft IQ: Work IQ, Foundry IQ, Fabric IQ and a new Web IQ — context as managed infrastructure.
Read this because A ~53x ARR multiple is a bet on agent-as-headcount, not agent-as-tool. The flywheel is the proof and the risk: Cognition writes ~90% of its own code with Devin, so its growth and its demo are the same thing — until growth slows.
Cognition, maker of the Devin coding agent, raised $1B+ at a $26B valuation (May 27) — 2.5x in 8 months, $492M ARR, ~90% of its own code AI-written.
Read this because Same week one model hunts vulnerabilities, Google ships one that auto-patches them. When attack and defense both run at machine speed, the patch window collapses from weeks to minutes — and the human moves from operator to auditor of agent-written fixes.
Google launched AI Threat Defense (May 27): a Gemini platform fusing Wiz, CodeMender, and Mandiant to find and auto-patch vulnerabilities at machine speed.
Read this because The bet isn't a model — it's the layer above all of them. OpenRouter sells optionality: route to whatever model is best/cheapest, lock into none. The tell is the cap table — Nvidia plus Snowflake, Databricks, MongoDB, ServiceNow all want a seat at the routing layer.
OpenRouter raised a $113M CapitalG-led Series B at ~$1.3B (May 26) — routing 100T tokens/month across 400+ models for 8M+ users.
Read this because The search bar is the new battleground. By embedding an action-taking agent into the box users already type in — and pushing it into Teams and ChatGPT — Salesforce bets the moat is CRM data and workflow context, not the chat UI. The interface is commoditizing; the data isn't.
Benioff unveiled Agentforce Coworker (May 21): an AI agent in every Salesforce search bar — taps live CRM data, takes action, spans Slack, Teams, ChatGPT.