Accountants are getting squeezed. The decline in new accounting graduates continues. At the same time, retirement of experienced accountants is accelerating. As these forces have collided, the accountant workforce has shrunk by more than 17% since 2020.
In order to meet the continued demands of the profession, accountants have increasingly turned to AI. Recent studies show that 74% use AI tools several times a week and 44% actually rely on those tools multiple times a day. Even more so, 34% have admitted to using “shadow AI”, or tools that are not officially approved by their firm.
Then on June 24, 2026, something notable happened. The IRS Office of Professional Responsibility (OPR) issued Bulletin 2026-19. It clearly stated “Practitioners must strictly handle all client data using only secure, enterprise-approved AI” and that “willful mishandling of taxpayer information through AI may lead to disciplinary actions under Circular 230.”
So where do things go from here?
Takeaway #1: AI Governance Is Critical
The bulletin makes clear that this is not only on the individual practitioner. Firms have responsibilities too. As the bulletin says, “Firms must deploy internal policies and procedures for compliance with Circular 230 in the AI space.” Staff must receive comprehensive training on how to use AI and how to evaluate its outputs. Firms must establish clear protocols for where AI can be used and what tools are approved for those use cases. Firms also need to carefully think about client consent since rules like IRC §7216 limit how tax return information can be used and disclosed.
It occasionally means some additional work to meet these governance requirements. But it protects client trust, which is the foundation this profession is built on. And it lays the foundation for the next era that this profession also likely will be built on.
Takeaway #2: Human-Centric AI Is Here to Stay
The bulletin also highlights the human role in all this. It calls AI a “powerful tool, not a substitute for professional judgment.” And it leaves no doubt about who owns the output: “Practitioners must thoroughly review all AI-created documents and language incorporated into writings before delivery to a client or submission to the IRS. Due diligence requires verifying the accuracy of facts, citations, and calculations produced by AI.”
So what does that mean for the people doing the work? As long as the rules keep a person on the hook for the call, the accountant has a job. But AI will likely reshape that job. As AI takes on the tedious work, an accountant's day may shift even more to the work that matters most. The judgment a hard call demands. The relationships built over years. The wisdom a client leans on when the stakes are high.
Takeaway #3: AI Will Likely be “Transformative”
Finally, the bulletin includes a very notable phrase. “Transformative potential” is used to describe what AI brings to tax practice. It also reminds practitioners that under Section 10.27(a), “A practitioner may not charge an unconscionable fee in connection with any matter before the Internal Revenue Service,” and that “cost savings should be passed on openly, with billing practices that reflect the efficiencies gained from the use of [AI].” Read together, the message is that AI fundamentally changes the economics of the work.
And professionals are thinking a similar way. Intuit's 2026 Firm of the Future report found that “77% agree the gap is widening between firms where AI is embedded in daily workflows and firms that use it only occasionally.” A profession that gets faster and cheaper tends to reset what clients expect from everyone in it. AI may create some big winners. AI may also create some unfortunate losers. However, the IRS makes clear that all firms must take a prudent approach when adopting AI.
Conclusion
So what is the bigger picture? There is risk in adopting AI. There is also risk in not adopting it. The market is being flooded with countless AI products and services. Most are increasingly capable of handling the tedious work. Fewer were built with Circular 230 in mind. That is the real fork in the road. When adopting AI, the key is to choose products and partners where compliance is built into the core of the offering, not tacked on as an afterthought.