What just happened? When KPMG, one of the world's biggest accounting firms, asked its own auditor for a discount, it wasn't because of sloppy books – it was because of AI. The firm told Grant Thornton UK that new technology should make audits faster and cheaper, and pressed for lower fees to match those efficiencies.
The negotiations, which took place last year, signal how deeply automation is beginning to influence professional services markets that have long been defined by human labor. For decades, audit pricing has reflected time, expertise, and regulatory complexity. Now, tools powered by machine learning and generative AI are changing that calculus.
People familiar with the talks told the Financial Times that KPMG International urged Grant Thornton to share any cost savings from its AI rollout across its audit operations. When Grant Thornton resisted, KPMG reportedly warned it might seek another auditor.
Grant Thornton had audited KPMG International – an umbrella organization coordinating the network's global operations – for several years. KPMG argued that its consolidated accounts were straightforward and that Grant Thornton's familiarity with the business should allow faster, more efficient work once automation tools were applied.
UK filings show the negotiations ended in KPMG's favor. Grant Thornton's fee for auditing KPMG International's 2025 accounts fell to $357,000 from $416,000 the previous year, a reduction of about 14 percent. Though both firms declined to discuss commercial terms, the outcome suggests AI may be starting to reshape not only how audits are performed but how they're priced.
Accounting firms have been pouring money into AI systems designed to improve accuracy and compress timelines. Machine learning tools now handle repetitive tasks like transaction matching and anomaly detection, while large language models assist with documentation and audit planning. KPMG has said its generative AI tools help auditors refine risk assessments, design substantive testing procedures, and document their findings more precisely.
Yet those innovations aren't necessarily making audits cheaper to run. KPMG International said that although AI can create efficiencies, developing and operating such systems also generates additional costs.
Audit economics remain in flux. Grant Thornton's UK audit leader Gary Jones wrote in a December blog post that the firm's work was becoming "faster" and "smarter" thanks to automation.
But he also noted that audit quality depends heavily on expert human judgment – a costly ingredient that AI doesn't replace. In a statement to the Financial Times, the firm said that fees "reflect both the cost of our people and the cost of the technology that supports them."
For other companies, KPMG's aggressive stance may set a precedent. Figures from Ideagen Audit Analytics showed that average audit fees across Europe continued to rise last year, with firms citing heavy investment in AI platforms. Only one country in the regional survey reported lower fees.
That trend highlights a growing tension: while AI is sold as a productivity multiplier, its transition costs can be high. Firms must build custom systems, train models on historical audit data, and integrate them with strict compliance workflows – all under regulatory oversight. The result is an uneven picture, where automation may boost speed without immediately cutting costs.
Still, KPMG's success in negotiating a discount from its own auditor gives weight to a question that many corporate clients are now starting to raise: if AI makes audits more efficient, when will the savings reach the customer?
KPMG asked its own auditor for a discount, citing AI "efficiencies"
