Deloitte agent rollout

Deloitte has begun putting artificial intelligence agents into its global audit platform, extending the tools to a workforce it says numbers about 85,000 auditors worldwide. (deloitte.com) The rollout centers on Omnia, Deloitte’s cloud audit system. On July 15, 2025, Deloitte said Omnia was adding generative artificial intelligence and “agentic” features that can review documentation, draft audit communications, extract data from documents, and help researchers answer accounting questions. (deloitte.com) Deloitte describes agentic artificial intelligence as software that can carry out multistep tasks with human oversight, rather than just answer prompts. In an August 24, 2025 post, the firm said these agents are meant to support auditors on repeatable workflows while people handle judgment-heavy work such as risk assessment and internal controls. (deloitte.com) Audit is a natural test case because large parts of the job already run through standardized checklists, document reviews, tie-outs, and control testing. Deloitte’s Omnia pages say the newest version is designed to automate some manual activities and give clients real-time visibility into audit progress. (deloitte.com) The scale matters because Deloitte is one of the world’s largest professional services networks. The firm reported more than 470,000 people globally in fiscal 2025, which means an 85,000-person audit population would make this one of the biggest disclosed deployments of agentic tools inside a professional-services workflow. (deloitte.com) Deloitte is also selling the same idea outside its own audit practice. Its July 2025 announcement said the firm had launched a Global Agentic Network, a catalog of ready-to-deploy agents that Deloitte practitioners can use to advise clients on large-scale deployments. (deloitte.com) The pitch from Deloitte is speed and consistency, not a fully autonomous audit. The firm says the tools perform first-pass reviews, create draft materials, surface risks from outside information, and connect across systems, while auditors remain responsible for oversight and conclusions. (deloitte.com) That framing also reflects the profession’s constraints. Audit firms face independence rules, documentation requirements, and inspection risk, so the safest place to deploy agents is inside tightly defined tasks where inputs, outputs, and review steps are already visible in software. (deloitte.com) Deloitte’s own language keeps returning to the same division of labor: agents for the repetitive work, humans for the judgment calls. In practice, that is where many large companies are testing artificial intelligence now — not as a general worker, but as a digital specialist inside a narrow, measurable process. (deloitte.com)