The impact of artificial intelligence on education has progressed exponentially. AI tools like ChatGPT are changing how students engage with coursework and how instructors like me must think about assessment, feedback and professional preparation.
Traditional exams and take-home assignments are no longer sufficient to evaluate true student understanding. As an accounting professor, I now find myself redesigning activities to ensure that students not only use tools like AI effectively, but also understand when not to rely on them, and how to add value beyond what a machine can produce.
Here are the most important lessons I have learned this academic year about integrating AI thoughtfully into accounting education:
AI can organize and assist — but it is not a verification tool
AI has proven useful in helping students organize their thoughts, draft outlines or simplify dense material. But I stress to my students that it is not — and cannot be — a verification tool.
Accounting requires precision and authority. When analyzing lease classifications, fund reporting or tax basis issues, students must turn to official sources: the FASB Accounting Standards Codification, the GASB Codification or the Internal Revenue Code. Using AI for clarity is fine; relying on it for correctness is not. Students must always check AI-generated content against the source materials our profession depends on.
Use AI to brainstorm — not to customize client advice
I also teach students to use AI as a brainstorming tool, particularly during the research or planning phases of a project. It is helpful for generating directions or outlining general concepts. But it is not appropriate for tailoring advice to a specific client or scenario.
Accounting is fact and circumstance driven. Whether advising nonprofits, preparing a tax return or planning an estate strategy, professional recommendations must be based on facts, rules and professional judgment. Client work should never feel generic or templated. AI cannot replicate the critical thinking and ethical responsibility that real-world client service demands. We may think it does, but I grade enough assignments to know it does not.
Completeness matters: Do not let AI gloss over critical details
A major challenge I have encountered is that students sometimes assume AI-generated answers are complete. But AI often skips over technical nuances, and that is a problem in accounting.
Take the example of gifted versus inherited property. If AI omits mention of the alternate valuation date, dual-basis rules for gifts, or the carryover implications for depreciable assets, the answer may sound right but dangerously incomplete. I now build assignments around the question: What is missing? This teaches students to verify, expand, and improve on what AI provides — a key workforce skill. The students do not just supply and answer; they explain how they got to the answer and why it is correct.
AI cannot replace the human touch — only you can do that
Lastly, I tell my students this: the human element is irreplaceable. AI may draft a technically accurate paragraph, but it cannot show empathy, ask clarifying questions or build trust with a client.
Clients know when they are receiving a canned response. They want professionals who listen, understand and deliver advice that feels specific and authentic. I challenge students to review AI-drafted memos and ask: Would this make a client feel heard? Would they trust this advice? Regardless of whether the answer is yes or no, more human insight is always welcome.
The rise of AI has pushed me to rethink how I assess student learning, encourage original thinking, and reinforce professional values. I no longer focus solely on whether students arrive at the correct answer, but on whether they can communicate clearly, verify sources, and apply judgment — the skills that truly matter in practice.
This year has reminded me that while technology is changing fast, our responsibility as educators remains the same: to prepare students to be thoughtful, ethical and human-centered professionals. AI may change how we teach, but it does not change why we teach.
