Featured

Workday Launches “Agent Passport” — The First Enterprise-Grade Safety Check for AI Agents

Why it matters: Workday — used by 65%+ of the Fortune 500 — just unveiled the first product designed to test, verify, and continuously monitor every AI agent touching enterprise HR, finance, and IT systems. This isn’t a white paper. It’s shipping with Cisco as the launch testing partner.

Agent Passport vets agents against the OWASP LLM Top 10, NIST AI RMF, and MITRE ATLAS standards — testing for prompt injection, jailbreak, goal hijacking, system prompt extraction, and employee data leaks. At runtime, it can allow, block, or route agent actions in real time. A single revocation can stop affected agents company-wide.

“AI agents are now doing the most sensitive work in the enterprise,” said Dean Arnold, VP AI Platform at Workday. “One insecure agent can leak employee data, break compliance, and put the company on the front page for the wrong reasons.”

The bigger picture: This is the governance layer enterprises have been waiting for — independent, auditable, and tied to public standards. Early access starts Q3 2026, GA before year end. If your company runs Workday and is experimenting with AI agents, this should be on your radar now.

🔗 Workday PR | CIO.com | InfoWorld

⚡ GitHub Copilot’s Pricing Overhaul Hits 4.7M Users — Agentic Bills Jump 10x–50x

GitHub Copilot switched to token-based billing on June 1, ending flat-rate pricing for 4.7 million paid users. Power users running agentic coding sessions are reporting projected cost increases of 10x to 50x — one Reddit user estimated their bill jumping from $50 to $3,000/month. Code completions remain free, but chat, agent mode, and Copilot CLI now draw from a monthly credit pool. Internal Microsoft documents obtained by journalist Ed Zitron revealed Copilot’s week-over-week operational costs had nearly doubled since January, making the switch urgent. Takeaway for business leaders: The era of “all-you-can-eat” AI tools is ending. Budget for agentic AI like you budget for cloud compute — by consumption, not by seat.

🔗 TechTimes

⚡ Survey of 132 Enterprise AI Leaders Reveals the “Governance Mirage”

VentureBeat’s Q1 2026 Pulse Research found that while 43% of enterprises claim a central team owns AI governance, 23% can’t agree on who holds ultimate responsibility. The survey identifies the “Governance Mirage” — org charts that look good on paper but lack actual technical control layers. The core finding: production failures aren’t caused by models — they’re caused by the runtime layer most teams patch with retries instead of fixing. As AI shifts from chatbots to autonomous agents that book flights and update ERP systems, the infrastructure gap is fueling “Shadow AI” — departments building ad-hoc agents without IT oversight. The fix: an “Agentic Fabric” that decouples intelligence from execution with state management, guardrails, and observability.

🔗 The AI Chronicle | VentureBeat

Trump Signs Executive Order for Voluntary AI Model Review — 30 Days Before Release

President Trump signed an executive order on June 2 establishing a voluntary framework for AI developers — including OpenAI, Google, and Anthropic — to give federal agencies up to 30 days of pre-release access to frontier models for national security testing. The order explicitly prohibits mandatory licensing or preclearance for AI model development, making participation optional. It also directs the Attorney General to prioritize criminal enforcement against those using AI agents for unauthorized computer access. The order represents a middle path — slimmer than the mandatory vetting the administration shelved last month, but a significant shift for an administration that has resisted AI regulation. National security agencies will develop classified benchmarks to assess AI model cyber capabilities within 60 days.

🔗 White House

ChatGPT Becomes First AI App to Hit 1 Billion Monthly Active Users

OpenAI’s ChatGPT crossed 1 billion monthly active users in May 2026, reaching the milestone in just three years — faster than TikTok (5 years) and Instagram (8 years). The milestone, reported by Reuters via market intelligence data, cements ChatGPT as the fastest-growing consumer app in history. Meanwhile, Anthropic’s Claude reached 56 million MAUs with 640% year-over-year growth, though U.S. users spent 5% less time on ChatGPT the following month — suggesting the AI app market is becoming genuinely competitive. Both OpenAI and Anthropic are racing toward IPOs: Anthropic filed confidentially this week following a $65 billion funding round at a nearly $1 trillion valuation. OpenAI is expected to announce its own IPO filing in the coming weeks.

🔗 U.S. News / Reuters | Times of India

Anthropic Scales Claude Mythos to 15+ Countries — 10,000+ Critical Flaws Found

Anthropic is expanding Project Glasswing — its zero-day vulnerability hunting program powered by Claude Mythos — to approximately 150 organizations across 15+ countries, including Australia, Canada, Japan, South Korea, and major EU nations. Since April, Mythos has helped partners surface over 10,000 high- or critical-severity flaws. New partners include Okta, Samsung, SK Hynix, SK Telecom, NATO, and the EU cybersecurity agency ENISA. Anthropic describes the target as any organization where “a successful attack on their codebase could be catastrophic” — estimating a single major breach could affect more than 100 million people. The expansion comes one day after Anthropic confidentially filed for IPO and OpenAI released its rival GPT-5.5-Cyber model.

🔗 TechCrunch

🚀 Want AI working for YOUR business? Most companies are experimenting with AI chatbots. We deploy AI workforces — AI Employees that follow up on leads, resolve support tickets, publish content, chase invoices, and screen 200 job applicants overnight so your hiring manager starts Monday with the top 10. Each role has a cost profile and human oversight, managed through one platform. This newsletter? Written by an AI Employee, approved by a human — so our team stays focused on what only humans can do. AIToken Labs helps businesses design their AI Workforce Operating Model — starting with the 2-3 roles that deliver ROI in the first 60 days. Book a free 40-minute AI Workforce Blueprint Session. →