EY has deployed enterprise-scale agentic AI within its global Assurance practice, covering 160.000 audit engagements and over 10.000 in the MENA region.
The deployment integrates a multi-agent framework built on Microsoft Azure, Microsoft Foundry, and Microsoft Fabric into EY Canvas, the firm's single global Assurance technology platform. EY Canvas currently processes more than 1.4 trillion lines of journal entry data per year and supports the daily workflows of 130.000 Assurance professionals. The integration follows an extended period of testing and piloting, and the firm expects the technology to support all end-to-end audit activities by 2028.
Technology underpinnings and strategic context
The agentic framework is designed to help audit teams orchestrate complex tasks and processes, address risks more dynamically, and access continuously updated auditing and accounting guidance. The modernised audit methodology aims to reduce administrative burden on clients, improve risk assessments, and preserve the role of professional judgement throughout the engagement lifecycle.
In addition, the deployment is part of what EY describes as a multibillion-dollar investment in audit quality, technology, and people, carried out under its 'All in' global strategy.
Alongside the technology roll-out, EY is implementing a structured global training programme designed to upskill all audit and technology risk professionals during 2026. The programme includes immersive and in-person learning components and is intended to be updated continuously in response to regulatory, technological, and methodological developments.
The firm is also introducing dedicated AI assurance services intended to support clients undergoing AI-led business transformations. These cover AI diagnostics, governance, risk management, and controls frameworks, enabling organisations to assess AI readiness, manage AI-related risks, and strengthen accountability structures.
The breadth of the deployment, spanning both the scale of engagements and the number of professionals involved, signals a wider shift within the audit profession towards AI-integrated workflows. As regulators and standard-setters continue to develop frameworks around AI use in financial reporting and assurance, the adoption of agentic AI by major audit networks is likely to draw scrutiny and, in time, shape emerging standards for the field.
