The U.S. Accounting Shortage Is So Severe That AI Can’t Replace Workers Fast Enough

Silicon Valley executives routinely forecast a future where artificial intelligence eliminates millions of white-collar jobs within months. The auditing profession is living in a starkly different reality: firms cannot find enough humans to do the work, and the shortage threatens the financial infrastructure that keeps global markets functioning.

DataSnipper CEO Vidya Peters estimates that open roles in audit and finance consistently outnumber available professionals by a factor of two to two-and-a-half. While technology leaders warn of mass displacement, the industry actually faces record attrition and a labor deficit so deep that automation serves as a survival mechanism rather than a cost-cutting weapon.

Federal employment data quantifies the collapse. The Bureau of Labor Statistics recorded roughly 1.96 million accountants and auditors working in the United States in 2019. By 2022, that figure had plummeted to approximately 1.65 million. Although the workforce partially recovered to about 1.77 million by 2025, the profession remains nearly 200,000 workers short of its pre-pandemic baseline.

Universities are not producing replacements fast enough to close the gap. The American Institute of Certified Public Accountants reported that schools awarded 55,152 accounting bachelor’s and master’s degrees during the 2023-24 academic year, a 6.6% decline from the previous year. Spring 2025 enrollment finally showed a 12.4% increase—the highest since 2020—but the demand side continues to accelerate. The BLS projects 115,300 annual openings for accountants and auditors through 2035, driven primarily by retirements and permanent exits from the field.

This scarcity fundamentally alters how artificial intelligence functions in regulated industries compared to sectors like copywriting or basic legal services, where labor supply was relatively balanced before generative models arrived. In fields governed by strict regulatory oversight, AI tools must operate under direct human supervision because regulators hold licensed professionals legally responsible for errors. An algorithm cannot sign an audit opinion, nor can it face enforcement action from the Securities and Exchange Commission if financial statements prove materially false.

Peters draws a sharp distinction between building automated agents and reviewing their output. Designing and approving an agent requires senior expertise comparable to managing a junior auditor. Evaluating an agent’s results, however, falls within the capabilities of far more staff members. Platforms built for this environment must enforce traceability, linking every generated figure directly back to source documents so reviewers can verify claims instantly.

Yet trust among practitioners remains volatile. DataSnipper’s research found that 78% of audit and finance professionals trusted AI in 2025, up from 74% in 2023. That confidence collapsed to 55% in the firm’s 2026 AI in Audit report, marking the sharpest decline in four years of tracking. While 73% of respondents consider the technology essential, only 13% of organizations have integrated it into actual workflows. Professionals operating under strict liability standards resist adopting tools they cannot independently verify.

The stakes extend far beyond corporate balance sheets. Auditors validate the security testing behind biometric authentication systems, confirm that taxpayer funds properly subsidized public transit infrastructure, and ensure retirement accounts reflect accurate valuations. When oversight fails, the consequences cascade through the economy. The 2001 collapse of Enron, initially questioned by Fortune journalist Bethany McLean months before the company filed for bankruptcy, directly triggered the Sarbanes-Oxley Act and reshaped federal securities law.

Amsterdam-based DataSnipper has structured its own growth around this tension between speed and accountability. Unlike AI startups raising tens of billions in primary venture capital, the company remained bootstrapped and profitable before securing a $100 million Series B led by Index Ventures in 2024 at a $1 billion valuation. That round consisted entirely of secondary shares, allowing founders and employees to sell personal equity rather than diluting the company further. Every employee holds ownership stakes.

Peters, who spent over a decade in product and marketing roles at Intuit, MuleSoft, and Marqeta before relocating to the Netherlands in 2020, was recruited directly by DataSnipper’s founders in June 2023 rather than installed by investors. Under her leadership, the firm shifted from shipping one or two product updates annually to releasing new versions every business day.

The customer base reflects the scale of the verification problem. DataSnipper serves the Big Four accounting networks alongside major firms including RSM, BDO, Baker Tilly, Aprio, Grant Thornton, and Crowe. Corporate clients using the platform for internal audit and finance operations include Barclays, Netflix, Mastercard, Morgan Stanley, and Hilton. Government entities such as the U.K. National Audit Office, the European Parliament, and the national governments of South Africa and Australia also rely on the software. The company now operates across more than 3,000 organizations in 175 countries, with its Amsterdam headquarters employing staff representing approximately 70 nationalities.

DataSnipper launched Alwin, its agentic automation platform, in late September 2026.