Burnett Says Maybe — If You Know the Rules.
By CPA Trendlines Research
Generative AI is no longer optional for tax professionals—but neither is skepticism.
Veteran tax attorney and educator Bradley Burnett is urging CPAs and tax professionals to engage with AI now, warning that delay carries its own risk. Don’t, he says, “get left behind at the train station,” he says.
LIVE WEBINAR: Is AI Fit for Tax? with Bradley Burnett, Mon, Dec. 29. Register | Learn more
Still, Burnett is unequivocal that AI must be handled carefully. He cautions that tools like ChatGPT amplify the user’s competence—or lack thereof. “ChatGPT in the hands of someone without proper experience and training is like handing a loaded gun to a monkey,” Burnett says, underscoring the professional liability risks of blind reliance.
A core principle in Burnett’s guidance is understanding what AI is—and is not. He describes generative AI as a scavenger rather than an expert. Generative AI is “a forager,” he says, meaning it can surface both high-quality insights and dangerously misleading information that looks authoritative on first read.
That risk is magnified by AI’s outdated knowledge base. Burnett notes that public models like ChatGPT acknowledge that much of their training data stops around September 2021. As a result, “it mostly misses the boat on giving relevant, currently accurate answers which rely upon new developments,” particularly for post-2021 tax legislation.
To mitigate risks, Burnett urges tax professionals to build disciplined AI workflows. Central to that discipline is verification. “Human validation and refinement are necessary,” he says, emphasizing that responsibility for accuracy never shifts from the practitioner to the software.
Burnett also encourages CPAs to enhance their interaction with AI by asking more informed questions, requesting citations, and verifying results across multiple platforms. He advocates “platform hopping,” moving between AI tools, Google searches, and authoritative research databases inside paywalls to confirm and update results.
When used properly, Burnett argues, AI delivers real productivity gains. “Generative AI takes the grind out of certain tasks,” he says, pointing to drafting client communications, summarizing long documents, organizing data, and kickstarting research memos as high-value use cases.
He likens effective AI use to collaboration inside a firm. “It’s a bit like walking down the hall and bantering with colleagues,” Burnett says, which may help frame issues and surface ideas, but is never a substitute for professional judgment.
Burnett’s bottom line is that AI proficiency itself is becoming a core professional skill. Those who develop it will have a competitive edge, while those who don’t may struggle to keep up. As he puts it: “AI won’t replace accountants, but an accountant with AI experience will replace other accountants without AI experience.”