What challenges did EY face before AI implementation?

Perform high-quality audits globally with 150+ countries and 130,000 auditors, managing massive data volumes and complex risk assessments. Reduce administrative burdens and improve process efficiencies while preserving the balance of AI support and human judgment in audits. Integrate AI capabilities holistically within the existing EY Canvas platform to support all audit phases, risk assessment, and fraud detection.

What was the impact of the AI solution?

Improved audit quality and process efficiency on a global scale across 150+ countries and 130,000 auditors. Data-driven dynamic risk management and operational audit orchestration reduce the burden on human auditors and accelerate audit cycles. Enhanced insights and confidence in audit results. Projected full AI integration into EY audits worldwide by 2028. Positive market impact with accelerated adoption of AI-driven audit enhancements.

Source

MicrosoftProductionEvidence: Low40/100

EY Enterprise-Scale Agentic AI for Audit with Microsoft Foundry, Fabric, and Azure

Use case typeAudit automationUpdated Jun 13, 2026

EY developed a multi-agent AI framework embedded in their EY Canvas audit platform using Microsoft Azure, Microsoft Foundry, and Microsoft Fabric. The AI system processes 1.4 trillion journal entry lines annually to enhance audit orchestration, dynamic risk management, and quality control across EY's global network with 130,000 auditors in 150+ countries. EY rolled out global auditor upskilling programs for effective AI adoption and expects full AI coverage of audit activities worldwide by 2028, improving audit quality, reducing administrative load, and maintaining human judgment.

Organization: EY
Industry: Professional Services
Location: United States
Published: April 2026

Reported outcomes

Strategic outcomes

Risk & complianceEnhanced fraud risk assessmentBetter decisions & insightDynamic risk managementSpeed & agilityStreamlined audit workflowsEmployee experienceGlobal auditor upskilling programs

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: EY
Provider: Microsoft
Maturity: Production
Linked source: CPA Practice Advisor

EY rolled out global auditor upskilling programs for effective AI adoption and expects full AI coverage of audit activities worldwide by 2028, improving audit quality, reducing administrative load, and maintaining human judgment

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 2 of 2

1Agentic AI
2Audit Automation

Perform high-quality audits globally with 150+ countries and 130,000 auditors, managing massive data volumes and complex risk assessments.
Reduce administrative burdens and improve process efficiencies while preserving the balance of AI support and human judgment in audits.
Integrate AI capabilities holistically within the existing EY Canvas platform to support all audit phases, risk assessment, and fraud detection.

Implemented a multi-agent AI framework on Microsoft Azure cloud incorporating Microsoft Foundry and Microsoft Fabric, tightly integrated into EY Canvas audit platform.
The AI analyzes extensive accounting data, dynamically assessing risk and orchestrating audit activities with data-driven insights.
The solution supports end-to-end audit phases including fraud risk assessment, enhancing quality and streamlining workflows.
Launched global upskilling programs for auditors to maximize AI adoption and benefit realization.

Technologies

Microsoft Azure Microsoft Foundry Microsoft Fabric EY Canvas

Improved audit quality and process efficiency on a global scale across 150+ countries and 130,000 auditors.
Data-driven dynamic risk management and operational audit orchestration reduce the burden on human auditors and accelerate audit cycles.
Enhanced insights and confidence in audit results.
Projected full AI integration into EY audits worldwide by 2028.
Positive market impact with accelerated adoption of AI-driven audit enhancements.

Sources & evidence1

Evidence: Low40/100Evidence strength