Auditor-in-Control Checkpoints: What EY UK's 2025 Audit Quality Report Says About AI, October 2025

EY UK's 2025 Audit Quality Report is dated October 28, 2025. It is the firm's report on its audit practice, and AI is one part of it. What follows reports parts of what the firm says there about its use of AI, as the firm's own claims, not weighed here.

Technology in the Audit

In 2025, the report says, EY "has introduced more than 30 new and enhanced capabilities across the Assurance technology platform," including "the integration of artificial intelligence at scale." It describes a tool that uses generative AI to search and summarize accounting and auditing content "with citations," delivered within the workflow and tailored to the context of each engagement. Its guided workflows, it says, "prompt structured consideration of cybersecurity, AI use, and the reliability of information produced by the entity."

On data, it states a principle in its own words. "A core principle of the EY Digital Audit is that the provenance and quality of underlying data are non-negotiable."

Tested, Approved and Released

"To ensure AI is used safely, ethically and in line with leading global standards and regulations, we have developed the EY Responsible AI framework," the report says. It describes the framework as principles-based, and as aligned with leading ethical frameworks and standards, including those outlined by NIST, the ISO, the OECD and the EU AI Act.

The firm then says how its AI capabilities are approved and released. "Our AI capabilities are risk-managed, tested and approved against principles in this framework, then released with clear guidance and auditor-in-control checkpoints. This approach ensures innovation is conducted with the discipline and traceability expected by stakeholders."

"Auditor-in-control checkpoints" is the firm's phrase, and the report does not say what one looks like. What would a reviewer find on an audit file to show that a checkpoint was passed, and by whom?

People, and What Comes Next

"Our people's experience and professional judgment remain critical to planning and executing a high-quality, technology-enabled audit," the report says, and the firm says it will continue to give its teams training and guidance on how to use its AI and other automation tools appropriately. It mentions "responsible use of AI training."

It says "the next evolution is agentic AI," which it describes as capabilities that can orchestrate multi-step activities across platforms, connect knowledge and analytics, and propose next actions, "always with human oversight and approval." Among its actions for FY26 is to further embed AI and technology in its audits, to respond to the risks present in the companies it audits, "as well as innovating our audit approach."