AI labs have been positioning their agents as ready for complex, long-horizon workflows. A new benchmark released today puts that claim to the test in one of the highest-stakes environments possible: U.S. healthcare. The results are not reassuring. actAVA.ai released CHI-Bench, described as the world’s first long-horizon healthcare benchmark for AI agents. Testing 30 frontier agents across 75 real-world U.S. healthcare workflows, the benchmark found that the best-performing agent — Claude Code with Opus 4.6 — achieved a 28% pass rate at pass@1. That means even the top performer fails approximately 7 out of 10 real clinical cases. ...

A year ago, your organization had a dozen AI agents. Today you have hundreds, maybe thousands. Every team moved fast, built their own setup, and now you have a sprawling estate of agents — each with different authentication schemes, different tool access, different logging configurations. Then the CISO asks: “Which agents are accessing customer PII?” That’s the problem Databricks addressed in a May 20 engineering blog post, and the solution they’re presenting is Unity Catalog extended to govern the entire agent stack — not just data, but models, tools, and the MCP servers that connect agents to enterprise systems. ...

Google I/O 2026 had a lot of announcements — the official recap lists over 100 of them. But two stand out as genuinely landmark developments for the agentic AI ecosystem: Universal Cart and Gemini Omni. Together, they signal that Google isn’t just building AI features — it’s building the infrastructure for AI agents to act autonomously in the real world. Universal Cart: Agentic Commerce, For Real Google has tried to build shopping features into Search for years. But Universal Cart is different from every previous attempt — because it’s architecturally agentic. ...

Jensen Huang didn’t mince words on Nvidia’s Q1 FY2027 earnings call. “Agentic AI has arrived,” he told investors — and the numbers backed him up. Nvidia posted $81.615 billion in revenue for the quarter, commonly rounded to $82 billion, representing 85% year-over-year growth driven by insatiable demand for AI infrastructure. The company also unveiled what may be its most consequential hardware announcement since the H100: the Vera CPU, Nvidia’s first chip purpose-built for agentic AI orchestration. ...

Your AI agent has a ticket queue full of infrastructure requests. It has read access to your runbooks, write access to your deployment pipelines, and the ability to execute changes against live systems. It also reads Jira tickets, wiki pages, and Slack transcripts to decide what to do next. That combination — broad access plus natural-language reasoning from untrusted inputs — is the attack surface security teams need to be thinking about right now. ...

In one of the most significant talent moves of 2026, Andrej Karpathy — OpenAI co-founder, former Tesla Director of AI, and one of the most influential AI researchers alive — has officially joined Anthropic’s pre-training team. Karpathy announced the move on X with characteristic directness: “Personal update: I’ve joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D.” ...

One of the biggest barriers to deploying AI agents in enterprise environments isn’t the model quality — it’s the plumbing. Specifically: how do you give a cloud-hosted AI agent access to your private internal APIs without opening your firewall? Anthropic just solved this with MCP Tunnels, announced yesterday at the Code with Claude London developer event. The Enterprise Firewall Problem Model Context Protocol (MCP) has become the standard way to give Claude agents access to external tools — think databases, ticketing systems, data warehouses, feature-flag services, internal APIs. The problem is that most enterprise infrastructure lives behind firewalls, private VPNs, or air-gapped networks by design. ...

Big pharma has officially gone agentic. Bristol Myers Squibb — one of the world’s largest pharmaceutical companies, responsible for life-saving drugs like Opdivo, Eliquis, and Revlimid — announced today that it has signed a strategic agreement with Anthropic to deploy Claude Enterprise as its shared intelligence platform across its entire global operations. The scale is significant: more than 30,000 employees worldwide will gain access to Claude’s advanced reasoning and agentic capabilities. This isn’t a pilot program. This is whole-company rollout. ...

Google didn’t just announce a faster model at I/O 2026. They announced a different kind of model. Gemini 3.5 Flash, unveiled on May 19, is explicitly positioned as agentic-first — the first model in Google’s lineup designed from the ground up not for chat, but for the kind of long-horizon, multi-step, tool-using work that defines modern AI agents. It’s fast enough for real-time loops, capable enough for complex reasoning, and it’s already the engine powering several of Google’s biggest product launches this week. ...

Google’s Antigravity started as an AI-powered coding tool. With the launch of Antigravity 2.0 at Google I/O 2026, it’s become something far more ambitious: a complete, standalone agent-first development platform with a new CLI, a Python SDK, and a Managed Agents API that plugs directly into the Gemini API. This is a ground-up rebuild, not an incremental update. What’s New in Antigravity 2.0 The Desktop Application Antigravity 2.0 ships a redesigned desktop application built specifically for multi-agent workflows. The headline features: ...