CFO Brew helps finance pros navigate their roles with insights into risk management, compliance, and strategy through our newsletter, virtual events, and digital guides.

CFO Brew

Technology

Courtney Vien

cfobrew.com

Provisions of the European Commission’s (EC) AI Act that affect general-purpose AIs are set to 

 this August. But leaders of both US and EU companies, and even some EU heads of state, have called the legislation premature and are asking for a pause.

The AI Act first went into effect this February, 

 that fall into its “unacceptable risk” category, such as social scoring systems and AIs that scrape the internet or CCTV for facial recognition data.

On August 2, “general-purpose” AIs, or those with wide uses that can be integrated into many different applications, will be required to comply with EU copyright law, disclose their technical documents, and provide the EC with a summary of their training data. The AI Act will be fully in effect, minus a few exceptions, by August 2026, according to the EC’s website.

But some companies and politicians have argued that much of the Act isn’t ready for prime time. As the Center for European Policy Analysis (CEPA) pointed out, some of the AI standards are 

, and some EU member states have yet to establish AI oversight bodies.

In an open letter to the EC, 46 leaders of large EU companies—including Airbus, Lufthansa, microchip lithography company ASML, and French AI startup Mistral—asked for a “

’” on the AI Act. Swedish prime minister Ulf Kristersson and other EU leaders have requested a pause as well.

So have some US companies, including Google and Meta. The Act will affect 

 that have clients or customers in the EU. Businesses with direct operations in the EU, whose AI might touch EU companies in their supply chains, or whose AI models process EU citizens’ data, should see 

The EC has stated that the AI Act will proceed as scheduled. “Let me be as clear as possible, there is no stop the clock. There is 

. There is no pause,” EC spokesperson Thomas Regnier said at a press conference, Reuters reported.

But Henna Virkkunen, the EC’s tech head, has said she’ll decide by the end of August whether a pause is needed. CEPA predicts that the February ban on unacceptably risky AI models will stay in place, but that regulatory obligations for high-risk and general-purpose AIs will be pushed out one to two years.

CFOs, along with their CIO and CISO colleagues, will definitely want to watch this space.

EU AI regulations go live next month

But companies and politicians have requested a pause.

Jesse Klein

You can probably now add AI training to the list of required work videos you aren’t allowed to fast-forward through. Buuut, maybe there aren’t even enough companies requiring AI instruction to skip.

 from OwlLabs and Pulse, almost 70% of companies have added AI tools to their workflows but just 38% are actually training their employees how to use it.

Companies without AI training risk investing in a tool no one uses (or uses poorly), creating a workforce that falls behind the competition. An improperly trained workforce also risks misuse of AI tools and consequently organizational 

“People in positions of power tend to roll this out thinking everybody will just start using it. But that’s simply not what happens,” said Conor Grennan, Chief AI architect at NYU Stern and CEO of AI training consultancy AI Mindset, . He has provided training to NASA, McKinsey, PwC, Amazon, Unilever, JPMorgan, and others.

But before signing everyone up, CFOs should understand that AI training isn’t just about teaching humans how to use a new tool, it’s about training them to think differently, to identify biases, and to change behaviors, according to Grennan.

“People in finance tend to want to have a new tool,” he said. “But this is not a new tool. This is a new way of thinking and working.”

Here’s how some experts are approaching AI training as a mindset shift rather than a technical skill.

 Companies often treat AI like a software program with a steep learning curve, when it is more akin to healthy eating: something that people might understand intellectually but which requires a behavioral change to actually be effective, Grennan said.

“People are treating this like a general digital transformation, when it has actually everything to do with change management,” he said.

He focuses AI trainings on shifting mindsets around approaching work rather than on “prompt engineering” and new features. He thinks the finance department is particularly susceptible to erroneously viewing AI as the next Excel.

And just like eating healthy, don’t fall for the snake oil salesmen, warns 

, an AI consultant and trainer. If someone is promises immediate returns on investment or guaranteed productivity increases from a few hours of AI training, it’s probably too good to be true.

, and according to its Chief AI Officer, Dan Priest “materially changing the business…does require training.” He finds microlearning training videos for employees, hosts prompting parties where expert and beginner prompters come together, and invites business school professors to lead formal trainings on AI for business strategy. His team has also created low-code or no-code sandboxes for employees to safely experiment with the tools.

Priest’s AI training philosophy is grounded in practicality. He focuses trainings on increasing employee confidence and comfort with using tools and expanding their thinking of how the tools can be used.

Hopper tailors training tasks to mirror the daily work of the employees he’s coaching. Before the training, Hopper will get a report from the hedge fund and private equity group he is working with. He then demonstrates how to recreate the report more quickly using AI.

“Where the technology is today, it has become the equivalent of an exoskeleton,” he said. “You’re still driving, but you have these additional powers and efficiencies and speed with what you do.”

Grennan emphasizes that using an AI enterprise tool that’s been vetted by senior leadership to ensure it has

Priest’s team at PwC has invested in what he calls “responsible AI” by restricting what datasets teams are allowed to upload, creating firewalls within the programs, and implementing appropriate admin controls.

Hopper always talks to senior leadership before his trainings to understand the company’s AI usage policy or helps them create one before rolling AI out to the entire company. If leadership isn’t on board, no matter how good the training is, the employees won’t be able to use the AI for their work.

“If I don’t have the buy-in from IT and the security folks and the compliance folks of the company, then the training is meaningless,” he said.

Company leadership is racing to add AI tools, but do employees know how to use them?

Successful AI usage at a company comes from behavioral change, not the latest features.

CFOB VE-6 Steps for Building an AI-Ready Finance Team video embed

Speakers

Questions?

AI is changing what finance teams need to succeed, making now the time to rethink your talent strategy. Join CFO Brew and Oracle NetSuite to learn how to prepare your team with the right skills, roles, and use cases for an AI-enabled future.

6 Steps for Building an AI-Ready Finance Team

Alex Zank

It’s ineffective to drive a vehicle with a brand-new set of tires that are misaligned. It also seems silly, then, to funnel more money into technology budgets without aligning those investments with business goals.

A vast majority (93%) of organizations are spending more on technology, yet just 27% said their tech is “fully aligned with business goals,” according to a Grant Thornton survey of more than 550 executives.

“Business leaders recognize the need to invest in technology to enhance the customer experience, improve operations, and drive profitability,” Nick Vellani, national managing principal of Grant Thornton’s technology modernization practice, said in the survey report. “But they are competing with their internal technology function’s priorities. These may include infrastructure and application cloud migration, cybersecurity, and data management—all of which are critical to ongoing operations but do not directly align to business priorities.”

 Respondents most commonly said their firms were increasing technology investments somewhere between 6%–10% this year. But there were some particularly big spenders in the survey cohort: Nearly one in 10 said their firms are upping tech investments by 20% or more.

Business leaders said their organizations plan to shovel the largest share of their tech spending (37%) into platforms such as ERPs and CRMs. Just over a quarter of tech spend (28%) is going to targeted third-party solutions, while 19% is earmarked for custom development with internal resources, and 16% is going toward custom development by third parties.

 Companies with misaligned technology investments often share common traits, Grant Thornton found. They may be smaller firms that are using outdated systems. They could have made big tech investments without first laying out their business needs. Or they may have purchased multiple systems that aren’t compatible with each other.

Firms shouldn’t sleep on their data quality, either. About a third of respondents said “their data needs improvement to support technology initiatives.” Just 16% said their firms had “excellent” data quality.

Tech spending and business goals are out of alignment for many orgs

Almost all business execs said they’re upping tech spend, yet only a fraction felt these investments are fully aligned with business goals.

The finance world might be in a state of high anxiety over AI and whether it’ll replace jobs. But for some of the accounting world’s leading lights, the technology’s giving them a case of deja vu.

Every time a new innovation comes out, pundits claim it will put accountants out of business, Mark Koziel, CEO of the AICPA, said during a keynote speech at that organization’s Engage conference.

And that’s been happening for a long time, he added. “The creation of the electronic calculator was the major innovation in the 1930s,” he said. “And you know what they said back in the ’30s? ‘Wow, with this electronic calculator, I don’t know why we’re going to need all these accountants anymore.’”

The same thing happened when the microcomputer debuted in the 1960s and the spreadsheet in the 1970s, Koziel noted. (In fact, demand for accountants and management analysts 

 after the advent of spreadsheets.) The advent of AI, he predicted to CFO Brew, means the profession will go through “a temporary blip of trying to figure out what AI” can do and where talent needs to be redeployed.

Tom Hood, EVP for business engagement and growth at the AICPA, likewise took a tempered, but still optimistic, view of AI.

“All the major CFOs we’re working with are saying they’re not thinking of AI as a headcount reduction [tool] in the immediate term,” he told CFO Brew. “In five years, they think they will be slowing up new hiring as they upskill their existing finance skills.”

But he also believes that AI will accelerate financial accounting’s transition toward “higher-value strategic roles,” he said. “And we’re, in fact, seeing that all over the place.” Newer workers will need to upskill quickly, he said, but will spend less time on the “soul-sucking, entry-level work that all accountants have to do.” They’re also “going to have to have the level of understanding how AI works to be able to understand whether its output is legitimate or not,” he said.

But Koziel isn’t that concerned about the profession’s ability to upskill. Accounting has had to do so multiple times in the past to keep up with technological change, he pointed out.

“What I did as an auditor 30 years ago, as a new associate, [and] what auditors are doing today, are night and day,” he noted. Before he became an auditor, success in that field was measured in terms of “being able to add up a trial balance within a short amount of time,” he said by way of example. But by the time he entered the field, auditors were uploading trial balances into Excel or audit platforms, and that skill didn’t matter so much anymore.

Accounting’s now facing another “learning gap,” Koziel said. Firms and academia will need to update their training to prepare new accountants for an AI-powered workforce, but the profession’s not “quite there yet.”

AI’s creating opportunities for accountants: 

AI comes with opportunities for the profession, Koziel said. Accounting firms are actually adopting AI more quickly than general businesses are, he pointed out, which means that firms can help clients implement AI.

And he thinks accountants could be called upon to provide AI oversight, in much the same way the profession provides SOC reports on companies’ data security practices. “We are the trusted profession for assurance over many things,” he said. “And I think AI is going to be a piece of that as well.”

The AICPA has also dipped its toe into the AI pool by building Josi, a chatbot solely for professional standards and guidance. “What’s cool is it’s a closed instance,” Hood said. “All of our data stays in there.” Unlike, say, ChatGPT, it doesn’t contain any outside data that could lead it to make the wrong inferences, he explained. Josi’s outputs come with source references, so CPAs can readily check it against the authoritative standards.

“That’s our beginning of doing something big with AI as an association,” Hood said.

The death of accounting has been greatly exaggerated

AI’s not going to do the profession in, AICPA CEO Mark Koziel said.

Accounting once ran on a whole lot of dead trees.

Jim Bourke, a partner at accounting firm Withum who’s been with the firm since 1987, remembers those days. Before the cloud, if you asked your accountant for a copy of an old tax return, for instance, “it could be a couple of weeks by the time you get it,” he told CFO Brew.

Staff would need to get it out of the physical file room at the office, photocopy it, and either mail it to you or have you pick it up, he said. (And cross your fingers that no one shelved your paperwork in the wrong file.)

The advent of cloud-based accounting software changed all that. The cloud “was a revolution,” according to Bourke, who, as managing director of his firm’s advisory services branch, has helped many firms build out their tech stacks.

SaaS made it possible for accountants to work remotely, to hire and find clients from pretty much anywhere with internet access, and to provide clients with faster and more convenient service. But getting firms on the cloud hasn’t all been smooth sailing.

Cloud software for accounting first emerged in the late 1990s with the advent of NetLedger, the ancestor of NetSuite. By 2009, accounting software giants CCH, Intuit, and Thomson Reuters all had web-based tax preparation programs. But the first generation of cloud products was “very feature-thin,” Randy Johnston, a consultant with 40 years’ accounting software experience, told CFO Brew.

By 2012, cloud technology had advanced to the point where the profession took notice, in the form of the inaugural Digital CPA Conference, originally named the Digital CPA 

, which took place in October of that year. Around 2013 was when cloud products “started to get enough features that they could kind of work,” Johnston said. One of the pioneers of cloud accounting, he sat in the original design meetings for QuickBooks Online, NetSuite, and Sage Intacct.

But accounting firms were still wary. Convincing them to give up their onsite servers “was like taking their baby away,” Bourke, an early adopter and cloud evangelist, told us. One partner at a firm told him, “‘I want to be able to put my arms around the server,’” he recalled. Firms were concerned about data breaches, but they also feared change in general.

“They got so used to the way in which they did things. Most firms went to the cloud kicking and screaming.” he said.

The pandemic was what really spurred firms to join the cloud, Bourke noted. When firms weren’t able to get into their offices or visit clients on-site, they quickly saw the value of being able to access data remotely. “Everyone finally realized, ‘Oh my God, this cloud thing makes a ton of sense,’” Bourke told us.

Desktop’s being sunsetted, but laggards remain: 

Today, many vendors have discontinued their legacy desktop products. Intuit stopped accepting new US subscriptions for QuickBooks Desktop in September 2024. Most new vendors entering the market only offer cloud products, Bourke said.

But not all firms are on the cloud yet. Some of the holdouts seem to be smaller firms whose clients haven’t made the switch: Johnston said he recently spoke with a small Texas firm whose 20 clients are all on desktop, and it hasn’t even considered moving to the cloud. Firms that convert now may “have to leave some of their client base behind,” he said. And converting can be costly, he added: “You have to dedicate the resources and charge enough to get it done, and sometimes the clients aren’t willing to pay.”

Bourke feels that getting on the cloud is a no-brainer. “It’s accessible anywhere day or night or in any country, anywhere in the world,” he said, and because it’s scalable, it works for sole practitioners and huge firms alike. It allows firms to give clients the “Amazon experience” they’ve come to expect even from accountants, he said, and provide them with more convenient, transparent, and faster service.

But there are some drawbacks to the cloud, according to Johnston. Cloud software often isn’t as robust as the legacy programs it replaced, he said. Then there’s SaaS price creep to contend with. Since switching software is a big undertaking, if vendors raise prices, firms may feel they have little choice but to pay up. “That type of stuff is rampant in the accounting market,” Johnston said.

Security is a legitimate concern, Bourke added, admitting, “Data breaches keep me up at night.” But firms can protect themselves by choosing good providers, he said. He recommends looking for vendors that have had SOC II audits to help ensure they’re using best practices for data security.

AI could be accounting’s next digital “revolution,” Bourke believes. He hopes it will catch on more quickly than the cloud did. Change, he said, is necessarily painful, but his advice is to “get it over with; put it behind you.”

“It’s where you’ve got to go for the profession to remain relevant for so many years to come,” he concluded.

The history of accounting’s cloud revolution

The long, slow road to the cloud was paved with dead trees. 

 is real, and it’s here. More than one in five professionals at tax accounting and audit firms said they’re already using generative AI technology, a big jump from 8% in 2024, a Thomson Reuters survey found. Just this week, accounting firm 

 a $1 billion investment in AI technology.

As AI gains a stronger foothold at accounting firms, the next logical question is how they’re using the technology, exactly. Executives at two of the six largest US accounting firms helped CFO Brew answer that question.

The big takeaway? AI will place employee productivity on a higher plane of existence.

“We expect AI to make us more productive, make sure people are focused on their highest and best use, and not spending their time on lower-value activities,” Eric Miles, CEO-elect of Baker Tilly, told CFO Brew. Baker Tilly, by the way, 

 with Moss Adams to become the sixth-largest accounting firm in the US.

 One of the potential promises of AI is that it’ll remove the mundane, drudge work from our jobs. For PwC CFO Colin Wittmer, that promise is panning out.

PwC teams have already created AI tools for account reconciliation and consolidations, payroll, expense reports, journal entries, and pieces of internal auditing, according to Wittmer.

“I’m starting with the low-hanging fruit, but you get really good results with the low-hanging fruit,” he said.

Wittmer told us one of the biggest improvements has been around the company’s monthly management report. At the end of the month, his team prepares an analysis of its three lines of service: assurance, tax, and advisory. According to Wittmer, “the AI has accelerated that by at least a week.”

Dashboards that pull and aggregate from databases are nothing new, but for Wittmer, it’s “the query-ability” that sets the AI apart. He can load in the monthly package and ask the AI where the company should focus if it wants to, say, improve the tax business’s profitability. And it spits out an answer.

PwC is also making a billion-dollar investment in AI, and recently 

 corporate LLM tools, including a proprietary ChatPwC tool using Microsoft Azure’s OpenAI Service that integrates with OpenAI’s technology. The company has also given 95% of its US workforce AI training, and nearly 80% of its employees use the AI tools daily, according to Wittmer.

To help implement the tech, a specialist team at the company works in two-week sprints to see if it can automate a task using ChatPwC or ChatGPT. The team trains the AI by querying to “train the model to do what a human otherwise would do. And there's a lot of monitoring...They get better with time. They get better with more queries,” he said.

“Innovation should happen everywhere. Industrialization of an innovation should happen at the center and then go back out,” he added.

 like addressing a user need, “integrates into existing workflows, and drives measurable

improvements in efficiency, accuracy, and decision-making,” then the automation is expanded for the entire organization.

For PwC’s management report, Wittmer concluded the AI was successful because he saw increased accuracy and consistency, and his team spent less time gathering the data. When it came to AR reconciliations, his confidence in rolling out the tool stemmed not just from its efficiency and “reduction in manual effort,” but from the fact that many on his team were already using and trusting it. (One KPI for PwC is strong user adoption.)

Baker Tilly also has an innovation team that considers how new AI tools could improve efficiencies and how easily they can be implemented, according to Miles. The “ultimate test” for that team, he said, is to pitch their proposed innovations to service line leaders who’d be using it in practice, to see if the tool “has legs.”

“If an AI proof of concept can help us serve the middle market better, that’s got to be a criterion,” Miles said. “If it’s just an efficiency enhancement, but doesn’t make us more effective with a client service, to me, that’s not as impactful.”

“It’s smart," PwC’s Wittmer said. “And you can train it. But human intuition is hard to mimic.” And sometimes the AI is missing important nuances that are unique to PwC’s business, he said.

He wants his finance team at PwC to move beyond grunt work and focus on the business’s most impactful areas—using that human intuition for pricing and market analysis, developing new services, and identifying ways to improve cash flow.

“Every CFO has a list of 50 projects that [if] they could get to it, they would do it, but their people are tied up in things that AI could actually basically do for them,” he said. “Let’s take those people and focus them on higher-impact stuff.”

And in terms of job replacement, Wittmer said he wouldn’t be hiring “50 people to work down [his] project list” anyway. Instead, he envisions a future where the firm grows while headcount remains stable.

This tracks with what other finance professionals told CFO Brew at the 2025 Gartner CFO & Finance Executive Conference. Leaders at 

 said that, while they’re not letting anyone go, they also didn’t plan to grow their finance teams much, either.

Jeff Ferro, CEO of Baker Tilly, said AI will “enhance” rather than outright replace the work of many accounting professionals. Job replacements will happen only at the “lowest levels,” he predicted.

Alternatively, other executives expect a significant impact to white collar workers—which could include those in finance. Anthropic CEO Dario Amodei 

 that AI could eliminate half of all entry-level office jobs in the next few years.

Regardless, the AI revolution is here, and public accounting seems ready to capitalize.

“It’s coming,” Ferro said of AI reforms. “What it’s actually, finally going to look like, I don’t think anybody really knows, but it is coming and it is going to disrupt.”