One thread ties together every major announcement this week: agentic AI is no longer a prototype. It is production software, deployed at enterprise scale, reshaping how the agentic AI enterprise workforce actually operates. EY just handed AI agents to 130,000 auditors. Canva rebuilt its entire platform around agentic workflows. And a critical security flaw in the protocol connecting these AI agents to your systems is sitting unpatched. If you manage people or technology, this is your Monday morning reading list.

EY Deploys Agentic AI Across 130,000 Auditors in 150+ Countries

EY announced a global rollout of enterprise-scale agentic AI across its entire Assurance workforce, 130,000 professionals conducting 160,000 audits in more than 150 countries. The multi-agent framework, built on Microsoft Azure, Foundry, and Fabric, is now embedded directly into EY Canvas, the firm’s unified audit platform that processes over 1.4 trillion lines of journal entry data per year. (Source: EY Global Newsroom)

This is not a pilot. It is a full production deployment backed by what EY calls a “multibillion-dollar” investment in audit quality and technology. The agentic AI enterprise workforce at EY will have AI handling risk assessments, tailoring workflows to specific engagements, and reducing administrative burden on clients. Full end-to-end AI-supported audits are expected by 2028. EY is also part of the inaugural class of Microsoft’s Frontier Firm AI Initiative, alongside 13 other organizations. (Source: CPA Practice Advisor)

For HR leaders, this is the clearest signal yet that agentic AI is moving from “interesting experiment” to “mandatory infrastructure.” If a Big Four firm is retraining 130,000 people to work alongside AI agents, ask yourself: what does your organization’s AI upskilling plan look like? EY is running a global training program throughout 2026, and they have also joined Stanford’s Human-Centered AI Industrial Affiliates Program. If your team lacks an AI skills development roadmap, you are already behind the curve.

What to do: Audit your own team’s AI readiness. Map which roles have repetitive analytical tasks that AI agents could handle. Start building training plans now, not after your competitors have already deployed. Companies already using AI agents for HR workflows will have a head start when agentic systems become the default.

Critical MCP Vulnerability Exposes 200,000 AI Agent Servers to Remote Takeover

OX Security researchers disclosed a systemic vulnerability in Anthropic’s Model Context Protocol (MCP), the industry standard for connecting AI agents to external tools and data sources. The flaw enables arbitrary command execution on any system running a vulnerable MCP implementation, giving attackers access to sensitive data, API keys, and internal databases. This is not a bug in one tool. It is baked into the official MCP SDKs across Python, TypeScript, Java, and Rust. (Source: The Register)

The scale is staggering: 150 million+ downloads, up to 200,000 vulnerable servers, and at least 10 high- and critical-severity CVEs across tools like Cursor, VS Code, Claude Code, and Gemini-CLI. OX Security’s team successfully compromised 9 of 11 MCP marketplaces during testing. Anthropic declined to fix the architecture, calling the behavior “expected” and pushing responsibility to developers. (Source: Infosecurity Magazine)

If your engineering team uses any AI coding assistant, or if your agentic AI enterprise workforce relies on MCP-connected tools, this affects you directly. The practical step: implement input validation on all MCP server configurations, restrict command execution permissions, and audit which MCP servers your team is running. Do not assume your AI tooling vendor has handled this.

Canva Launches AI 2.0: From Design Tool to Agentic Creative Workforce Platform

Canva unveiled AI 2.0 at its Create 2026 event, the company’s biggest product overhaul since launching in 2013. The platform now supports conversational design from natural language prompts, agentic orchestration across its design engine, a Memory Library that retains brand preferences across sessions, and connectors to Slack, Notion, Zoom, Gmail, and Google Calendar. Canva claims its proprietary models are 7x faster and 30x cheaper than comparable frontier alternatives. The update is rolling out to its first 1 million users now, with 265 million monthly active users total. (Source: Fortune)

For marketing and HR teams that use Canva for employer branding, social content, and internal communications, this changes the workflow. Instead of switching between tools, you describe what you need and the AI orchestrates it. A recruiter could say “create a job posting graphic in our brand style and post it to LinkedIn” and the system handles the rest. This is the agentic AI enterprise workforce pattern applied to creative production: AI agents that do multi-step work, not just answer questions. (Source: SiliconANGLE)

If your team uses Canva at scale, request early access to AI 2.0 and test it against your existing content workflows. The Memory Library alone, which learns your brand guidelines automatically, could save hours of back-and-forth with designers. Check out our guide to the top AI tools for HR for more options.

Quick Hits

• OpenAI launches GPT-Rosalind for life sciences. A specialized reasoning model for drug discovery and genomics, already in use by Amgen, Moderna, and Thermo Fisher. In RNA sequence prediction trials, its best submissions ranked above the 95th percentile of human experts. (VentureBeat)
• Adobe unveils Firefly AI Assistant. Orchestrates complex workflows across Photoshop, Premiere, and other Creative Cloud apps using conversational prompts, integrating 30+ AI models. Another creative platform going agentic. (AI and News)
• Japan’s AI Promotion Act shapes innovation-first regulation. Enacted in May 2025, Japan’s principles-based AI law imposes no penalties on businesses, relying on cooperation and transparency guidelines rather than the EU’s prescriptive approach. (IAPP)

The shift toward an agentic AI enterprise workforce is accelerating across every industry, from professional services to creative tools to life sciences. Whether you are an HR leader planning upskilling programs or a founder evaluating your AI stack, the question is no longer “should we adopt AI agents?” but “how fast can we deploy them safely?” For a deeper look at how AI is transforming human resource management, explore Asanify’s latest research.

FAQ

What is agentic AI in the enterprise workforce?

Agentic AI refers to AI systems that take independent, multi-step actions rather than just responding to prompts. In enterprise settings, these agents handle tasks like audit risk assessment, design orchestration, and IT service management autonomously, working alongside employees to complete workflows end-to-end.

How does the MCP vulnerability affect companies using AI agents?

The MCP vulnerability allows attackers to execute arbitrary commands on servers running Model Context Protocol implementations. Companies using AI coding assistants like Cursor, VS Code with AI extensions, or Claude Code should audit their MCP server configurations and implement strict input validation immediately.

Should HR teams prepare for agentic AI deployment now?

Yes. EY’s deployment of AI agents to 130,000 auditors shows that large-scale agentic rollouts are already happening in professional services. HR leaders should start mapping roles with repetitive analytical tasks, building AI literacy training programs, and evaluating which workflows could benefit from agent-assisted automation.

Not to be considered as tax, legal, financial or HR advice. Regulations change over time so please consult a lawyer, accountant  or Labour Law  expert for specific guidance.