AI tools reshape accounting workflows & client work

Artificial intelligence tools have moved from experimentation to routine use across parts of the accounting profession, with practitioners citing time savings in administration, changes in client work and growing attention on autonomous "agentic" systems.

Software suppliers have added AI functions to common workplace products. Microsoft has integrated Copilot into Microsoft 365. Vendors, including Xero and Oracle NetSuite, have also added AI features to their software.

Evidence and uptake

Survey findings from "Human + AI in Accounting: Early Evidence from the Field" point to differences between users and non-users of AI. The research cited an increase in general ledger granularity of 12% for users. It reported 4 hours saved per week on data entry. It reported eight days saved at the month-end. It reported higher billable hours of 21%. It reported 55% more clients supported per week.

Some practitioners said expectations set at the start of the year did not match what the technology delivered.

"If you look back even just 12 months, the expectations around AI sounded pretty implausible," said Kayur Patel, Director of genAI, PwC New Zealand. "However, what we've seen over 2025 is that not only has it met those expectations, it's exceeded them beyond even what those of us who work with the technology every day anticipated."

Patel also pointed to uneven adoption. He said smaller companies tended to move earlier because they could make decisions faster and had fewer internal hurdles for new tools. He said larger practices often faced more rules around integration and governance.

Everyday tools

Accountants have increasingly used AI in routine tasks rather than in high-profile transformation projects, according to Heather Smith of Anise Consulting. She described "AI-powered notetakers" as a leading use case across practices.

"It's allowing us to build a deeper understanding of our clients and it really frees your mind up so you can focus on what's being said, and not on the mechanics of taking notes and then trying to remember what was spoken about," said Heather Smith, FCA, Anise Consulting

Smith also described the growing use of public AI services such as ChatGPT and Google Gemini for drafting and refining written communication. She said accountants used large language models for letters to clients, tone changes and procedural documents.

"Language models are all about words," she said. "So, an accountant might draft a client letter and then put it through the model to soften the tone, put niceties around it and just make it more personable and human."

Smith also raised concerns about the risks of entering sensitive information into public models. She described security and privacy as drivers for the use of tools with tighter controls. She cited Microsoft Copilot as a commonly adopted option in firms.

Agents and workflow

While large language models have dominated the current wave of workplace experimentation, practitioners and advisers have increasingly focused on agentic AI. The approach uses software agents that take actions and complete tasks with less direct prompting than typical chat interfaces.

Tristan Tan of IFM Investors said accounting professionals needed to plan for these systems as they reach business processes.

"It's really monumental for businesses when you get a tool that can actually reason and take action," he says.

Tan said agentic tools would change workflows in areas where professionals face knowledge gaps or complex, unfamiliar questions. He said professionals would supervise AI agents that conduct research and propose outputs for review.

"You're going to get more generalists: people who know how to supervise the machine, manage outcomes and control the end product to be what it should look like," says Tan. "AI is forcing people to rethink their processes, identify inefficiencies and fix them."

Tan also said AI would shape how teams collaborate across experience levels. He said older staff would contribute deeper subject-matter expertise, while younger staff would bring comfort with digital tools. He argued that experienced staff often had stronger habits around testing technology outputs.

"It's all about how you understand and engage with technology," he says. "In some ways, the older, more experienced generations are in a better place with AI because they've grown up having to figure out and troubleshoot how to get technology to work.

"Meanwhile, the younger generation grew up with state-of-the-art, working technology. They are more inclined to accept outcomes and assume technology is always correct, without deeply interrogating what it is actually telling you, a fundamental requirement and skillset for the future."

Outlook themes

Technology watchers also point to a wider set of trends around AI and security. Gartner's list of top technology trends for 2026 included multi-agent systems and domain-specific language models. It also highlighted confidential computing, digital provenance, AI security platforms and pre-emptive cybersecurity. The list also included "geopatriation", which Gartner described as moving workloads from global hyperscale clouds to sovereign or local environments.

For accounting firms, the next phase of adoption looks set to involve closer scrutiny of data handling, governance and workforce skills as automation moves beyond drafting text and into systems that take actions across workflows.