The unicorn CEO betting her industry’s AI crisis isn’t the one you think
· Fortune
Most of the fear about AI and knowledge work is that there won’t be enough jobs. Vidya Peters has the opposite problem, and she thinks almost nobody covering AI is asking about it.
Visit bettingx.club for more information.
“This is a profession that is actually in crisis mode,” said Peters, CEO of DataSnipper, an Amsterdam-based audit and finance automation company that reached unicorn status in 2024. In an interview with Fortune last week, days after DataSnipper launched Alwin, its agentic automation platform, Peters was asked what she made of the current AI discourse. She answered with a staffing problem. “At any given time, there are two to two and a half x as many jobs open than there are people to do those jobs,” she said. The profession has the highest attrition rates, she added. It’s “very cyclical, like it’s got these huge peaks, and it’s really hard to sustain a role.”
For two years, Peters has been making the case in public that audit faces a crisis of scarcity. The general AI-jobs conversation has mostly missed it. Her argument pushes against a much louder story from Silicon Valley, where AI leaders keep warning that white-collar work is about to be automated. Most recently, Microsoft AI CEO Mustafa Suleyman predicted in February that most tasks involving “sitting down at a computer,” accounting among them, will be fully automated within a year to 18 months. Peters has a line for that. “I’m sorry, but you can’t expect a great brand if that’s your headline,” she said. “And they said it 18 months ago, and here we are.”
She also offered a warning, aimed at herself as much as anyone. “Whenever you ask this question, pay attention to who’s answering the question,” she said. Everyone talking about AI is “speaking of AI from their own biased, vested interest perspective.” She says her own bias is the customers she serves: audit and finance professionals, “where nobody thanks them for doing the job quickly… but they’re definitely held accountable if they get it wrong.”
The numbers behind the crisis
Federal data shows how deep the hole got. About 1.96 million Americans worked as accountants and auditors in 2019, according to annual averages from the Bureau of Labor Statistics’ Current Population Survey. By 2022, that number had fallen to about 1.65 million, a loss of more than 300,000 workers. The Wall Street Journal reported the exodus at the time as a 17% decline in two years. The workforce has since partly recovered, to about 1.77 million in 2025, but that still leaves the profession roughly 200,000 people short of where it stood before the pandemic.
The pipeline hasn’t refilled it. Schools awarded 55,152 accounting bachelor’s and master’s degrees in the 2023-24 academic year, down 6.6% from the year before, according to the American Institute of Certified Public Accountants. There are early signs of a turn: enrollment in accounting programs rose 12.4% in spring 2025, the highest total since 2020. But the demand side keeps climbing. The BLS projects about 115,300 openings for accountants and auditors every year through 2035, most of them to replace workers who retire or leave the field.
Hannah Seal, a partner at Index Ventures, which led DataSnipper’s Series B, wrote last September that audit faces what Peters calls “an untenable equation: growing demand and shrinking supply.” As regulations increase and businesses become more complex, Seal wrote, fewer people are joining the profession, “and those who do often burn out under the heavy manual workload.”
The invisible profession
Peters’s pitch to people outside the industry starts with the claim that we all depend on auditors more than we know. Unlock your phone with Face ID, she said, and an auditor has already checked the security testing behind it. Swipe into the subway, and a government auditor has confirmed that the taxpayer money behind the turnstile built infrastructure “and not just lining someone else’s wallet.” She walked through the rest of an ordinary morning the same way: your email, your 401(k), your heating bill. Behind each is “someone in an IT audit,” “the financial audit,” “someone in a utility audit.” “We assume things work,” she said, but “nothing works unless someone is checking and ensuring that you can trust that it works.”
When an audit does fail, she said, the whole country finds out. Her go-to example is Enron, whose collapse led to the Sarbanes-Oxley Act. Fortune readers know that story. Bethany McLean’s March 5, 2001, piece in this magazine, “Is Enron Overpriced?”, was among the first public questions about the company’s books, nine months before it collapsed.
In other white-collar fields, like legal services or copywriting, AI is hitting labor markets that were roughly in balance. Peters argues that audit AI is arriving in a labor market that was already broken. “AI can be a huge boon on the limited talent that’s actually in the job,” she said, helping “people who do more with less.” In her keynote at DataSnipper’s Connect 2025 conference in New York last year, she said AI is making audit rewarding again.
‘Everyone can be a human in the loop’
If Peters is right, the industry can’t treat AI oversight as a specialist job for its most technical people. Almost any working auditor has to be able to do it.
She separates two roles. Building and approving an agent should be tightly restricted. DataSnipper lets customers give that authority only to experienced auditors who understand the process and can review an agent “like you would review a junior auditor.” Reviewing an agent’s output, and deciding whether to trust it, is a different job, and one within reach of far more people. “You may not have to design the agent,” Peters said. “Everyone can be a human in the loop.”
She sketched a hypothetical to show how that works. Picture a junior auditor with a year or two of experience. On DataSnipper’s platform, his manager can see whether he has actually opened an agent’s work, left comments and signed off, or just let it run unchecked. “That is a really important job,” she said. It matters because, as she put it, “the regulator is not going to care what tech tools you use. They’re going to want to know that a certified auditor… has reviewed it and approved.”
The complication in her own company’s data
DataSnipper’s own research complicates the picture. In its 2025 survey of audit and finance professionals, 78% said they trusted AI, up from 74% in 2023. This year, according to the company’s 2026 AI in Audit report, “trust fell sharply in 2026” to 55%, “the most dramatic shift” in four years of the survey. The same report found that 73% of respondents call AI essential, but only 13% of organizations have integrated it into real workflows.
Peters’s answer is traceability. Professionals won’t hand work to a tool they can’t check, so Alwin ties every output back to a source document, called Snips. “If we say it’s 101 euros and 23 cents, you can mouse over it and you can see exactly where we got that number from,” she said.
“It’s great, I’ll take the efficiency of AI,” she said. “But what’s really, really important to me is that the people we serve can stand behind the work of the AI and are actually controlling what the AI does.”
Swimming against the tide
Peters did not train as an auditor. She finished high school in Salt Lake City, studied industrial engineering in college in Chicago, where she met her Dutch husband, and then spent 12 years in the Bay Area. She held product and marketing roles at Intuit until 2015, then served as chief marketing officer of MuleSoft, which went public in 2017, and chief operating officer of Marqeta, which went public in 2021. When she and her husband married, he made her promise to raise their children near his family. “I really thought, oh, let’s just move to California for a few years and surely he’ll change his mind,” she said. He didn’t. The family moved to Amsterdam in 2020, and for more than two years she kept working California hours for Marqeta. As she put it, “Friday is a working day that ends at midnight.”
DataSnipper’s founders brought her in as CEO themselves in June 2023. She pointed out, without being asked, that it “wasn’t an investor decision or a board decision,” which is rare. She says she still talks to the founders almost every week.
The company she runs is its own argument about how to build in the AI era. While AI labs raise money in rounds measured in the tens of billions, DataSnipper has grown on its own revenue. When Insight Partners made a minority investment in 2022, the company described itself as “bootstrapped and profitable since its founding.” Its $100 million Series B, led by Index Ventures in 2024 at a $1 billion valuation, made it a unicorn. Peters says DataSnipper has never raised primary capital: that round was secondary, with founders and employees selling some of their own shares. Everyone at the company owns equity. Peters said letting staff cash in was the point: European tech is less fluent in the “language” of equity upside than the Bay Area, and she wanted employees to feel it.
“We’re very proud to be profitable and have financially funded our own growth,” she said.
That hasn’t slowed the company down. DataSnipper counts the Big Four and “almost every large audit firm network in the world” as customers, including RSM, BDO, Baker Tilly, Aprio, Grant Thornton and Crowe. It also sells to the internal audit and finance teams of Fortune 500 companies such as Barclays, Netflix, Mastercard, Morgan Stanley and Hilton, and to public bodies including the U.K. National Audit Office, the European Parliament and the governments of South Africa and Australia. It has customers in over 3,000 organizations across 175 countries, and Peters says its Amsterdam office holds about 70 nationalities. When Peters joined, the company shipped one or two new versions of its product a year. Now it ships every workday.
Peters draws a straight line from how she runs the company to how she wants AI to work for the people who use it. Both come down to not letting speed outrun control. “I could have taken over from the founders and leveraged the heck out of it and raised more primary capital,” she said. “But again, it’s been really important to me to strike the balance between growth and accountability.” What she doesn’t want, she said, is “to wake up in the morning thinking, here are a group of auditors that actually are being pulled out by the SEC because the agent went rogue and exposed data.”
She opened DataSnipper’s recent customer conference with a simpler message: what you do matters. Afterward, several auditors came up and asked her to speak to their staff, because they’d forgotten why they do the work. “There’s dignity in doing great work, solid work that people trust and people rely on,” she said. “And societies are built on these jobs.”
For this story, Fortune journalists used generative AI as a research tool. An editor verified the accuracy of the information before publishing.
This story was originally featured on Fortune.com