When an accounting firm hands first-year tasks to AI, its leaders should keep some of those tasks with new hires as training. Each kept task needs a named manager who reviews it. The firm carries the cost in staff and manager hours.
Squire is an audit, tax and advisory firm with offices across the Mountain West. On October 6, 2026, it announced that three of its leaders had won awards from the Utah Association of Certified Public Accountants (UACPA). Two still lead the firm, and one has retired.
Squire's chief executive won an award for leadership and service to the profession. The announcement says she stayed committed to developing the people behind the firm's growth. Another honoree once ran a firm that Squire later acquired, and while leading it he mentored "numerous employees into partner roles". These honors are for developing people. A firm can keep developing people after AI takes over first-year tasks, if new hires keep some of those tasks.
A Stanford study points to fewer young workers in the jobs most exposed to AI. Three Stanford researchers used payroll records from ADP, a payroll company, covering millions of US workers. The latest version of their paper, dated August 2026, uses data up to June 2026. The researchers ranked jobs by how much of their work AI language tools could automate. They then split the jobs into five groups, called fifths, from least to most exposed.
In the two most exposed fifths, employment of workers aged 22 to 25 fell about 11% from November 2022 to June 2026. In the three least exposed fifths, it grew about 10%. Experienced workers showed no comparable gap between the most and least exposed jobs. The gap for young workers came mainly from hiring fewer of them, rather than letting more of them go.
Accountants sit in the most exposed fifth. In October 2022, they were the fourth-largest job in that fifth in the payroll data. The Stanford study also looked at jobs where AI mostly helps workers rather than replacing their tasks. There, employment was flat or rising, especially for experienced workers. The authors call their results early, descriptive signals rather than proof that AI caused the gap.
The payroll data leans toward larger firms. When the authors checked national surveys, the gap was smaller there than in the payroll data. The gap also shrinks once the authors account for workers' education. Some of the gap began before AI language tools spread.
The American Institute of Certified Public Accountants (AICPA) asked public accounting firms about hiring for its 2025 Trends report. Of the firms that took part, 75% expected to hire at least as many new graduates in 2025 as in 2024. But too few firms answered for the report to give a reliable estimate of graduate hiring across all public accounting firms.
A firm that moves first-year tasks to AI should keep some of those tasks with new hires, done by hand, as practice. One example is a first-year tax associate who prepares a few simple returns in full before checking returns drafted by AI. A named manager reviews each return and explains the corrections. The cost is the associate's and the manager's hours, and the firm carries that cost itself.
What to do first:
- Managing partner: list the first-year tasks in audit and tax that AI tools now do, before the next hiring plan is set.
- Head of each service line, such as audit or tax: pick the tasks each new hire still does by hand, and name a reviewing manager.
- Head of HR: each year, record how many new hires each service line takes on and how many are still there two years later.
