Canadian businesses and individual taxpayers are increasingly utilizing general-purpose artificial intelligence (AI) tools like ChatGPT for bookkeeping and tax inquiries, a trend that financial professionals warn could lead to significant errors and potential conflicts with the Canada Revenue Agency (CRA). A recent survey conducted by Dext, a bookkeeping software company, indicated that in 2025, 76 percent of accountants and bookkeepers observed a surge in clients relying on large language models for tax and bookkeeping advice, often resulting in frequent mistakes.
The survey, which included responses from 500 accountants and bookkeepers across Canada, highlighted common errors such as misinterpretation of business expenses (44 percent), incorrect tax claims or charges (43 percent), faulty personal tax planning (36 percent), payroll errors (35 percent), and erroneous business tax planning advice (35 percent). These mistakes can lead to productivity losses for accountants as they address avoidable errors and can result in extra costs for business owners and individual taxpayers.
The implications of these inaccuracies extend beyond wasted time and money. Nearly a quarter of respondents (27 percent) noted an increased risk of insolvency or business failure, while others pointed to a higher likelihood of misuse of AI outputs to support inappropriate or fraudulent claims (42 percent), escalating fines and penalties (40 percent), and heightened CRA scrutiny due to incorrect or late filings (38 percent).
Melissa Robertson, principal of research and thought leadership at CPA Canada, emphasized the risks associated with using generative AI tools for tasks like calculation and automation. She cautioned that while these tools can provide valuable assistance, their outputs should be closely reviewed, particularly as transaction volumes increase. “There needs to be checkpoints to validate what the tool is doing, followed by a review of what’s actually been completed,” she stated.
Robertson further pointed out that the risk of error is universal across various accounting areas. “I think there’s a lot of overpromise in what some AI tools can do right now, and there is a lot of risk when organizations just take those tools at face value,” she said. Ryan Minor, director of tax at CPA Canada, echoed this sentiment, noting that while AI can be advantageous for document retrieval, it may not always provide the most current information reflecting the CRA’s guidelines. He urged users to consult original documents before acting on AI-generated advice.
Minor warned that individuals unfamiliar with the industry might find themselves misled by AI outputs. “If you’re not in the industry and you don’t have a hunch what the answer would be, you may be misled,” he cautioned. Jason Heath, managing director of Objective Financial Partners, shared that he often receives client inquiries featuring copy-and-paste information generated by AI tools. While such information is frequently mostly accurate, it can contain gaps or fail to align with the specific circumstances of the client, occasionally even conflating U.S. and Canadian tax regulations.
The allure of instant access to free advice is particularly appealing given the notable costs associated with legal, accounting, and financial consultations. However, if users fail to ask the right questions or do not adequately prompt the tools, the results can be misleading. Heath cited common tax errors related to sales taxes on short-term rentals, such as those listed on Airbnb, where mistakes can accrue significant costs depending on the property type, often going unnoticed for years.
“The CRA isn’t going to provide leniency because you relied on AI advice that you thought was correct, nor would it if it were professional tax advice that was incorrect,” he cautioned. Heath urged individuals to regard AI as a research tool rather than a replacement for professional guidance. “I’d be hesitant to rely on AI,” he said, “but I would be supportive of using it to further your knowledge and help you ask the right questions.”
The increasing dependency on AI in the financial sector underscores the necessity for vigilance and thoroughness in review processes. As businesses and individuals navigate the complexities of tax regulations, the potential for risk remains significant, necessitating a balance between technological assistance and expert oversight.
See also
AI Tools Transform Software Development: GitHub Copilot Enhances Collaboration and Automation
AI Tools Boost Publication Rates by 300%, But Narrow Research Focus, Study Reveals
Herzog Launches RailSentry AI System, Achieves Zero Incidents at Hazardous Crossing
UH Manoa Integrates Google and Gemini AI to Transform Language Teaching Strategies
Claude Code Achieves $1B ARR, Revolutionizing AI Programming with Agentic Development
