Understand this story
The English summary is being updated. See the reporting timeline below.
Latest development · Report
KPMG’s “Project Exponential” Uses More Than 100 AI Agents to Conduct a Parallel AuditDevelopments
1 confirmed developments
- Source:The Decoder
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Nov 18
The Decoder
KPMG’s “Project Exponential” Uses More Than 100 AI Agents to Conduct a Parallel Audit
KPMG deployed more than 100 AI agents in the audit of a FTSE 250 company, working alongside the traditional team through every stage. The agents analyzed some issues in greater depth, but the formal audit opinion was still produced through the conventional audit process. The agents can collect, analyze, and document data and draft conclusions, but human intervention is needed when matching regulations with company documents. KPMG plans to expand the project to as many as 20 large audits, while the FRC has emphasized that auditors remain responsible for the use of AI, its outputs, and the supporting evidence.
Trending on Hacker News
KPMG Experiment Shows AI Agents Sometimes Ask More In-Depth Questions Than Accountants
KPMG UK used more than 100 AI agents to review the financial statements of an unnamed FTSE 250 company. In some cases, they investigated more deeply and raised questions that the human audit team had not asked. The agents helped gather information, analyze and record data, draft conclusions, and recommend next steps, but auditors still had to intervene after one regulatory requirement was incorrectly matched. The experiment was not part of a formal statutory audit, and KPMG plans to expand it to as many as 20 larger audits.
AI Trends