Earmark CPE

Earn CPE Anytime, Anywhere

AI

Why Most Accounting AI Will Hit an Auditability Wall

· ·

Every day, another AI agent promises to revolutionize accounting. But there’s a fundamental problem most tech companies don’t understand: AI accounting will hit what FloQast CEO Mike Whitmire calls “the auditability wall.”

While Silicon Valley churns out press releases about AI agents that can handle complex accounting tasks, a reality check awaits. In this episode of the Earmark Podcast, host Blake Oliver sits down with Mike Whitmire, founder and CEO of FloQast, to explore why accounting AI is fundamentally different from AI in other business functions. Rather than getting swept up in the AI marketing frenzy, FloQast stepped back to solve the core problem: how to harness AI’s power while maintaining complete audit trails and human oversight.

As Whitmire warns, “A series of companies will come out with AI agents that can do a lot of this work fairly accurately. Then they hit this auditability wall, and it creates a big problem for companies trying to scale.”

The Auditability Problem That’s Breaking Accounting AI

Unlike other business functions where AI mistakes can be shrugged off, accounting operates under rules most tech companies don’t understand. When a sales AI messes up a lead, the stakes are minimal. But in accounting, every transaction must be traceable, every decision documented, and every process capable of withstanding regulatory scrutiny.

This creates a fundamental conflict between how most AI systems work and what accounting requires. “AI is really about automating work, and agents are doing non-deterministic work,” Whitmire explains. “So that becomes a little scary when you’re thinking about auditability.” Most AI systems function as “black boxes.” They can produce results, but they can’t explain their decision-making process in the detailed, step-by-step manner that auditors and regulators demand.

The problem is about to hit the industry hard. When AI systems can’t provide proper documentation and audit trails, auditors are forced to recreate all the work, defeating the entire purpose of automation.

Rather than getting swept up in the AI marketing that dominates press releases from major ERP vendors, FloQast took a different approach. “We tried to avoid the noise and think about how AI should be applied to accounting,” Whitmire says. They started with their experience as former auditors and accountants, asking: How do you combine AI automation with traditional software code and human oversight to create something that actually works?

The answer required rethinking the entire approach to accounting AI, leading to a solution that preserves audit trails and human oversight while still delivering efficiency gains.

FloQast Transform: Building AI Auditors Can Actually Trust

Rather than chase the latest AI trends, FloQast built something different: an AI system that auditors can actually work with. The FloQast Transform product harnesses AI’s power while maintaining the audit trails financial reporting demands.

The approach is simple: let accountants describe their processes in plain English, then use that narrative to generate automated scripts and complete audit documentation. “You build your agents,” Whitmire explains. “You chat with the product and explain your process in pretty extreme detail.”

As accountants describe their workflow step by step, the system populates what looks like a familiar Excel workbook. “This Excel workbook will ultimately be the audit evidence,” Whitmire notes. This isn’t just a user interface choice. It’s a deliberate design decision to ensure every AI-driven process produces the documentation auditors expect to see.

Take FloQast’s benefit allocation journal entry example. The process starts with integrating with UKG Payroll to pull down employee data. The accountant describes each step: “integrate with UKG,” then “pull down information around names, dollar amounts, state,” then “populate column A with this, populate column B with this, and bold and make the header gray because that’s the format I like.”

The system combines different types of automation. For routine tasks, it generates deterministic code that produces consistent results every time. But when the AI encounters something new, like when FloQast hired its first Kentucky employee, it doesn’t guess. Instead, “it surfaces the question to the reviewer of that work,” Whitmire explains. The accountant can approve the change, and going forward, Kentucky will be handled properly.

This approach changes the accountant’s role. Instead of being the preparer who manually processes transactions, they become the reviewer who oversees AI agents and approves exceptions. “Our goal is to empower accountants to automate the really repetitive, rote part of this job. Elevate them into the reviewer of the more complicated work that the agent’s now doing,” Whitmire says.

The system preserves every prompt sent to the AI, every output generated, and every decision made. When auditors come knocking, they can trace exactly how each transaction was processed and where humans intervened. It’s the kind of comprehensive audit trail that makes regulatory compliance possible while still delivering efficiency gains.

Beyond transaction processing, FloQast applies AI to other areas like variance analysis. When account balances trigger materiality thresholds, the system analyzes the biggest transactions causing the change and drafts explanations. “It’s like balance went up because of boom, boom, boom, boom, boom,” Whitmire says. “It’s not these wonderful essays on how things change. It’s like a list of transactions.”

The Future of Accounting: Cyborgs, Not Replacements

The auditability challenge yields a surprising conclusion: rather than replacing accountants, AI will transform their role in ways that could solve the profession’s biggest problems. But this transformation requires rethinking what it means to be an accountant.

Whitmire envisions accountants becoming “accounting transformation information managers,” professionals who combine accounting knowledge with software engineering capabilities. “It will be much more like the merging of an accountant with a software engineer,” he explains. “So you have the accounting knowledge, supplemented by software engineering tools like FloQast, where they can take their accounting knowledge, use our product, and automate their work.”

This isn’t just about learning new software. It’s about fundamentally changing the structure of accounting work. Instead of spending hours manually processing transactions, accountants would deploy AI agents to handle routine work so they can focus on reviewing exceptions, making judgment calls, and ensuring compliance.

The career implications depend on where you are professionally. For younger professionals, Whitmire recommends “Get really good at technology, learn these tools as they come out, and continue to learn about accounting. You’re going to be a very, very valuable employee going forward.” For experienced professionals, “You need to be really great at reviewing the work. Continue to be really great leaders, and run great organizations.”

This evolution could address the profession’s talent shortage. By making accountants more productive and the work more intellectually engaging, AI could help attract and retain talent. “My hope is that it does a really good job of plugging the talent gap we talk about so much,” Whitmire notes.

But there’s a learning concern. Whitmire worries about newer professionals who might skip foundational manual work and jump straight to reviewing AI-generated results. “I feel like the old man saying this, but I did learn a lot doing the work manually and struggling through it,” he admits. He recalls learning about jet lease accounting by struggling through contracts and GAAP guidance—work that an AI could now handle instantly.

The solution may require restructuring how accountants learn their craft. Perhaps starting in accounting roles where they do manual work before moving into audit, rather than the current model, where most start as auditors reviewing work they’ve never performed.

As Oliver puts it, “I would rather manage AIs than manage people.” It reflects both the appeal and reality of this AI-augmented future. Managing AI agents eliminates many interpersonal challenges while allowing professionals to focus on technical and analytical work.

The accounting profession is heading toward becoming a hybrid of human judgment and AI automation. The question is whether professionals and firms will adapt quickly enough to thrive.

Regulatory Changes on the Horizon

The discussion also touched on significant regulatory changes that could reshape the profession. There are efforts in Congress to eliminate the Public Company Accounting Oversight Board (PCAOB) and transfer its responsibilities back to the SEC without additional funding, effectively ending independent audit oversight.

“When I was at EY, we were always scared of a PCAOB audit. So it was a thing that drove behavior,” Whitmire reflects. The fear-based incentive improved audit quality, even if the overall effectiveness is debatable.

Without the PCAOB, the industry would likely return to peer review, where accounting firms review each other’s work. As Oliver notes, “You’re not so afraid of your buddies reviewing your work.” That’s the same dynamic that led to audit failures before the Sarbanes-Oxley Act.

This regulatory uncertainty adds another layer of complexity to the AI transformation. Firms implementing AI systems need to consider current audit requirements and how oversight might change in the coming years.

The Path Forward: Auditability as Competitive Advantage

The accounting profession’s rigid requirement for auditability is often seen as a weakness. But it may become its greatest competitive advantage in the AI revolution. While tech companies rush to market with AI agents that promise to automate everything, firms that understand and embrace the auditability challenge will build sustainable, scalable solutions.

FloQast Transform demonstrates that the future isn’t about choosing between human judgment and AI automation. It’s about creating systems where they work together seamlessly. By preserving audit trails, maintaining human oversight for exceptions, and generating documentation that auditors can use, they’ve solved the fundamental problem that will likely sink many AI accounting startups.

For accounting professionals, this is a career evolution opportunity. The future belongs to those who combine accounting expertise with technology capabilities. These professionals will be empowered by AI to focus on higher-level analysis, judgment calls, and strategic work. The professionals who master these systems now will find themselves in increasingly valuable positions as the technology matures.

To hear the complete conversation about FloQast’s approach to accounting AI, including detailed technical examples and Whitmire’s predictions for the profession’s future, listen to the full episode above.

The Math Is Brutal: Every CPA Must Triple Their Productivity by 2035 or Face Professional Extinction

· ·

“When you chart out demand versus supply of people over time, what that math tells you is that ten years from now, 2035, every CPA in the profession will have to be 2.7 times more productive on a revenue per employee basis than they are today. That is crazy.”

David Wurtzbacher shared this projection on a recent episode of the Earmark Podcast. As the founder and CEO of Ascend, a private equity-backed platform that’s completed over three dozen firm acquisitions in just over two years, Wurtzbacher offers an outsider’s perspective on the profession.

His background scaling Lightwave Dental from 7 to 80 locations taught him how private equity can either destroy professional cultures or transform them for the better. Now he’s applying those lessons to accounting, where the numbers paint a sobering picture: demand for services keeps climbing while fewer people enter the profession each year.

To put this in perspective, a typical well-performing firm today generates around $200,000 in revenue per employee. Wurtzbacher’s projection means that number needs to approach $600,000 per person within a decade. Even scarier? By 2035, roughly 85% of the profession will consist of people with ten years or less of experience in an industry where most say you can’t even make partner in that timeframe.

But Wurtzbacher isn’t just highlighting the problem. Through Ascend’s model of preserving firm independence while providing enterprise-scale resources, he’s showing how firms can achieve these seemingly impossible productivity gains through three key transformations.

The Leadership Evolution: From Managing Partner to True CEO

The biggest barrier to 2.7x productivity isn’t technology or talent. It’s how firm leaders spend their time. Most managing partners remain trapped doing client work while trying to run their businesses, creating a fundamental ceiling on growth.

“The very first place we go is to the leader of the firm,” Wurtzbacher explains. “We want to help them through a transition to become a true CEO, defined as them having one client, which is the firm.”

This leadership trap stems from what Wurtzbacher calls the “fiercely independent” culture of accounting. During his research, he consistently heard from entrepreneurial CPAs who valued their independence: the name on the door, community reputation, caring for people and clients their way. But this independence prevents the changes necessary for breakthrough growth.

The problem runs deeper than time management. The client service orientation that defines quality accounting actually caps leadership development. With seasonal demands and constant client pressure, managing partners find limited windows for strategic work throughout the year.

The real breakthrough requires confronting a limiting belief. “When you’re close with your clients, you believe nobody can do the work but you,” Wurtzbacher observes. “No one else can have this client relationship.”

Consider Lee Cohen from LMC in New York, who exemplifies this transformation. Cohen was initially stressed, unhappy, and heavily involved in client work. Through Ascend’s CEO transition process, “Cohen literally became a different person. He would tell you that,” Wurtzbacher says.

Fifty percent of Cohen’s transformation came from a mindset shift. The other fifty percent came from bringing in a Chief Growth Officer—not a traditional business development role, but a general manager from outside the profession. “A lot of them have MBAs, but they are hungry, humble, smart people that come in and create visibility for that leader about what’s going on in the business and where there are opportunities.”

This operational support, combined with the mindset shift away from client dependency, sets leaders free to focus on what only they can do: building and directing their firms.

Creating an “Irresistible Offer” for Top Talent

Even the best leadership transformation can’t solve the profession’s talent crisis through traditional methods. When quality candidates routinely field six, seven, or eight job offers, firms need something fundamentally different.

Wurtzbacher’s solution centers on creating an “irresistible offer,” and it starts with better recruiting. “So many firm recruiters grew up in the profession, and they’re trapped with the baggage of old ways of doing things,” he explains. Ascend built a team of professional recruiters from outside accounting who understand best practices for finding candidates and closing deals.

But the real breakthrough is compensation innovation. While the profession is “very base salary heavy,” Ascend developed an off-the-shelf bonus program that lets firms pay more cash than competitors. They also extended equity ownership far beyond traditional partner levels.

“We have well over 100 people across all our firms that are managers or senior managers that are investors in Ascend. They own Ascend stock,” Wurtzbacher reveals. These employees invest $10,000 to $50,000 annually in company stock—typically funded through the enhanced bonus program—essentially dollar-cost averaging into equity appreciation throughout their careers.

This creates what Wurtzbacher calls “a different cultural energy.” When people understand how equity value creation works outside the traditional partnership model, they connect their daily work to long-term wealth building. The psychological shift from employee to owner fundamentally changes commitment levels.

The design also solves a collaboration problem. Because everyone owns Ascend stock regardless of which firm they work for, “it creates a one team attitude across all our firms” that unlocks knowledge sharing across the platform.

The results speak for themselves. Firms that described capacity as their “#1 issue” now consider that problem solved. “Our big issue now is how do we go and get all the right kinds of new business that we want to keep our great people excited and motivated,” Wurtzbacher notes.

Technology at Enterprise Scale

Achieving nearly triple productivity requires more than incremental improvements. It demands systematic transformation through AI, global teams, and automation that individual firms cannot afford alone.

But there’s a gap between AI hype and reality. “There is so much more hype and future forecasting than there is reality in this area,” Wurtzbacher observes. For firms feeling behind, “that’s just not the case.” Most firms implementing AI are saving perhaps two hours per person per week, and that’s only for the most advanced adopters.

This creates both opportunity and strategic imperative. While individual firms struggle with overwhelming AI options, they lack technical expertise and capital for truly transformative capabilities. The solution requires enterprise scale.

Ascend illustrates this advantage in action. They’re building a 30-person software engineering and AI team by year-end. “No medium-sized or smaller firm is going to be able to do that,” Wurtzbacher explains.

Their strategy operates on two fronts: strategic buying versus building. For general needs, they purchase existing products. For capabilities essential to their workflows, they invest millions annually developing proprietary AI solutions.

One promising area addresses what Wurtzbacher calls the client context problem. Years of relationships generate institutional knowledge typically trapped “in your head, in spreadsheets, in work papers, in your inbox, and some other tool.” Their AI team works on aggregating this context into accessible systems that transform practitioners from information gatherers into true advisors.

Global talent represents another productivity component. Ascend’s acquisition and transformation of Sentient Solutions, a global capability center exclusively serving US accounting firms in Hyderabad, India, demonstrates sophisticated global team integration. But this isn’t simple outsourcing; it requires developing playbooks that elevate rather than replace domestic work.

Even basic infrastructure offers huge opportunities. Practice management systems in accounting are “so messed up,” Wurtzbacher notes. Before AI delivers transformation, firms need fundamental technological foundations for tracking work and maintaining institutional knowledge.

The Choice Facing Every Firm

Survival depends on three interconnected transformations happening simultaneously: leaders evolving from client servers to strategic CEOs, revolutionary talent approaches through equity ownership, and enterprise-scale technology investments individual firms cannot achieve.

This is a watershed moment for professional services. The mathematical reality of 2.7x productivity gains will separate surviving firms from those becoming obsolete. When 85% of the profession will have a decade or less experience by 2035, traditional models don’t just fail; they become mathematically impossible.

But there’s reason for optimism. Firms embracing these changes discover that freeing leaders from client work unleashes strategic energy, equity ownership creates cultural transformation beyond salary increases, and enterprise-scale technology delivers impossible productivity gains.

Wurtzbacher’s personal timeline reinforces this long-term vision. At 37, he tells people “this very well could be the last thing I do. So I’m thinking of Ascend in terms of decades.” While typical private equity investments last three to four years, his commitment spans the time needed for real transformation.

For accounting professionals, this is an existential threat and an unprecedented opportunity. The mathematical moment of truth has arrived. The question isn’t whether change is coming. It’s whether you’ll lead it or be overwhelmed by it.

Listen to the full conversation with David Wurtzbacher on the Earmark Podcast to hear more about Ascend’s approach to transforming accounting firms while preserving their independence.

Inside QuickBooks Online’s Biggest Transformation Since Going Cloud-Based

· ·

You’re reviewing a client’s profit and loss report when you notice little sparkle icons next to several expense categories. Curious, you hover over one and get an instant explanation: “Office supplies increased 127% compared to last month due to these three transactions.” What used to require detective work across multiple screens now happens automatically, with AI explaining not just what happened, but why.

This isn’t a future vision—it’s happening right now in QuickBooks Online’s July 2025 updates. On the latest episode of The Unofficial QuickBooks Accountants Podcast, hosts Alicia Katz Pollock from Royalwise and Dan DeLong from School of Bookkeeping break down Intuit’s massive “In the Know” session, where the company unveiled what they’re calling “QuickBooks on the Intuit platform.”

The transformation goes far beyond typical software updates. AI agents now work like digital detectives, scouring your data for patterns and anomalies. Banking feeds can automatically process PDF statements. Client communication occurs directly within QuickBooks, eliminating the spreadsheet shuffle. And those sparkle icons on reports? They’re AI-powered insights flagging unusual trends before your clients notice them.

But here’s what every accounting professional needs to understand: this isn’t an optional upgrade. By September 2025, everyone will be permanently on the new platform, with no opt-out option. The window to influence the final product closes soon.

AI Agents Become Your Digital Workforce

The heart of QuickBooks’ transformation lies in what Intuit calls “Agentic AI”—intelligent agents that actively hunt through your data for insights. Alicia explains her mental image: “I always imagine an AI bot in a detective hat, because that’s how I think about the AI is looking through the data and scouring it.”

The accounting agent, available for Essentials plans and higher, represents the biggest shift in how bookkeepers handle transactions. Instead of facing a wall of uncategorized entries, the system now identifies transactions that are “data-backed and likely to be accurate” and pre-checks them for posting. When three transactions meet this criterion, a banner appears announcing “three transactions ready to post.”

The game-changer is anomaly detection. Those sparkle icons appearing next to categories on profit and loss reports identify unusual trends automatically. Dan shares his experience: “I’ve seen it on some reports where the prior month there was a specific project that was done, and it said it right there on the screen like it went down this amount of percent because these two invoices were in the prior month.”

The categorization intelligence has evolved beyond simple pattern matching. The AI now recognizes that Shell and Arco are both gas stations, suggesting similar categories across different vendors. It scrapes bank descriptions for contextual clues and provides multiple suggestions for ambiguous transactions—offering both “meals and entertainment” and “travel meals” for restaurant charges, depending on your patterns.

Perhaps most significantly, categorization history has expanded from 12 to 24 months—a change Alicia specifically requested. This ensures annual charges can reference the previous year’s categorization, eliminating frustration with recurring yearly expenses.

Platform Integration Changes Everything

What Intuit calls “QuickBooks on the Intuit platform” represents more than rebranding—it’s the breakdown of decades-old product silos. As Dan explains, “their core offerings of TurboTax, MailChimp, and QuickBooks are getting homogenized here. And they can essentially talk to each other.”

The logic makes sense when you consider user patterns. As Alicia notes, “a lot of people use MailChimp who have never used QuickBooks. There’s a lot of people who file their taxes with TurboTax who have never used QuickBooks. So merging them all together is a natural evolution.”

The new interface features an app carousel with customer hubs, sales hubs, accounting hubs, marketing hubs, and business tax hubs. The customer hub will integrate MailChimp directly within QuickBooks, while business tax functionality brings TurboTax capabilities to the accounting workflow.

The enhanced bank feeds represent the most visible daily change. Alicia, who has been beta testing and providing daily feedback to developers, describes the evolution: “Everything that we knew and loved about the banking feeds is still there, but they kind of changed it.” The new system allows inline transaction editing, customizable column displays, and comprehensive transaction details.

The revolutionary statement import feature can process PDF bank statements and extract transactions automatically. While currently requiring human oversight—hence the two-hour processing time, at least for now—this capability could eliminate entire businesses built around transaction import services. As Alicia explains, “there’s a human being looking at it to see if it did a good job or not, and if it didn’t do it right, it’s actually going to a human being who is fixing the programming.”

Interface changes aren’t just cosmetic. The new left navigation is “brighter, it’s lighter, it’s prettier” with collapsible sections and bookmark functionality for one-click access to frequently used screens. The transformation from “Add” to “Post” in banking feeds reflects more technically accurate accounting language.

Client Communication Gets Built-In

The context gathering system eliminates the bookkeeper’s perpetual question: “What was this transaction for?” Built directly into QuickBooks, this feature threatens third-party apps by providing client communication tools within the core platform.

Alicia explains the problem this solves: “When you don’t know what something’s for, you have to go ask. And in the old days, we used to use spreadsheets for that. More recently, we’ve been using apps like Uncat, Keeper, or Financial Cents, where you can communicate with your clients right inside the app, but now you can do it right inside QBO.”

The system creates a to-do list maintained within QuickBooks, allowing bookkeepers to ask clients questions without requiring client QBO access. Clients receive emails with magic links to respond, and “it’s always the same link. And so you can just have your clients save it and bookmark it as the place to go.”

The expense forwarding feature allows anyone to send not just expenses but also income transaction directly into the system. However, this convenience introduces new risks. Alicia warns, “If you don’t have a bill approval process, you may have somebody who just goes in and pays everything without questioning anything. You actually could wind up paying bad actors who just sent random bills into your account to see if they could.” She reminds everyone to make sure they only give these email addresses to people they can trust.

The integration of Bill Pay Basic across all plans, including Simple Start, amplifies these concerns. Firms handling bill payments may want to consider upgrading clients to QBO Advanced, which includes mandatory bill approval workflows.

The September Deadline and What It Means

The timeline carries strategic implications beyond software preference. This isn’t a typical update where holdouts can postpone adoption—it’s a mandatory migration with a hard September deadline.

July offered opt-in/opt-out flexibility. August brought automatic transitions for new brand files. Crucially, all ProAdvisors’ clients were switched simultaneously. As Dan notes, “They threw accountants a bone” by ensuring firms wouldn’t juggle clients across different interfaces. September completes the mandatory transition, and by the month’s end, the new platform becomes permanent with no opt-out option.

The current period is critical for shaping the final product. As Alicia emphasizes from her beta testing: “This is the time to make sure that the platform works for us. They need your feedback.” Her daily communication with development teams resulted in interface improvements that serve real accounting workflows.

For firms considering the timeline, the choice is clear: engage now to influence the outcome, or adapt in September to whatever system emerges. The difference between being a beta participant and a forced adopter could determine whether your practice thrives or struggles.

Training and Resources Coming

Recognizing the scope of change, Intuit announced new training opportunities. Two courses are coming in October: one about understanding Agentic AI in general, and another specifically about AI agents in QuickBooks. There’s also ongoing research about what accounting professionals want to see in ProAdvisor Academy.

Alicia is completely rebuilding her training library at Royalwise. “I’ve got over 50 different courses of over 100 hours of QuickBooks Online content. So in September we are going to start over again from scratch,” she explains. Her Community and Coaching memberships will provide free entry into all webinars as she recreates content for the new platform.

Shape the Future or Be Shaped by It

The July 2025 QuickBooks updates represent the most significant transformation since moving to the cloud. AI agents are becoming the invisible workforce handling pattern recognition and routine categorization. New communication tools eliminate constant client back-and-forth. Interface changes reflect a fundamental shift toward integrated business management.

For accounting professionals, these changes represent both opportunity and risk. Those who engage now can influence the final product through feedback. As Alicia’s daily communication with developers shows, active participants can achieve solutions that serve the profession’s real needs.

But come September’s mandatory transition, the window for input closes. Firms will adapt to whatever system emerges from this beta period. The most successful professionals will view this transition as evolution—an opportunity to eliminate tedious data entry and focus on high-value advisory work.

Don’t let this transformation happen to you—be part of shaping it. The September deadline isn’t just about software—it’s about the future of the accounting profession itself.

Alicia Katz Pollock’s Royalwise OWLS (On-Demand Web-based Learning Solutions) is the industry’s premier portal for top-notch QuickBooks Online training with CPE for accounting firms, bookkeepers, and small business owners. Visit Royalwise OWLS, where learning QBO is a HOOT!

AI Won’t Just Speed Up Your Close – It Will Eliminate It Completely

· ·

“I want it gone.”

Aaron Harris, CTO of Sage, isn’t talking about making the financial close faster. He wants to eliminate it completely. No more monthly scrambles to lock the books. No more accountants working late to reconcile accounts. No more rigid cycles that control how businesses operate.

He shared this goal during his recent appearance on The Accounting Podcast, recorded at Sage Future in Atlanta. Harris has been a returning guest since 2019, and his message has stayed remarkably consistent: artificial intelligence will fundamentally change accounting processes and how businesses operate.

Harris isn’t just talking about automation making things faster. He’s challenging the basic business cycles that have defined corporate operations for generations. He envisions a future where annual audits become continuous, where quarterly tax filings disappear into real-time government systems, and where rigid business cycles give way to always-on, intelligent operations.

From Simple Tasks to Autonomous Operations: The Three Waves of AI

Harris breaks down AI’s evolution in accounting into three distinct waves, each building toward his vision of eliminating business cycles completely.

Wave One: Task-Based AI

The first wave focused on very specific jobs like reading invoices or classifying transactions. These systems worked like sophisticated scripts. They could automate tasks, but they needed humans at every step. “You can’t really interact with this AI,” Harris explains, “and because these are sort of very narrowly defined models, they can’t do a lot very flexibly.”

Wave Two: Generative AI

This wave brought conversational interfaces like Sage Copilot. Suddenly, AI could interact naturally with users and work more flexibly. This opened up possibilities for people outside the accounting team to use these systems. “The two big things are now you can interact with the AI,” Harris notes, “and it’s those underlying capabilities allowing that interaction that allow the AI to work more flexibly.”

Wave Three: Agentic AI

This is where Harris sees the real transformation. These systems can plan, execute, and operate on their own. They can access tools and interact with other systems without constant human guidance. “The real breakthrough comes with Agentic AI, where we’re now equipping these large language models. They think through how to plan something start to finish and execute on that.”

The progress has been dramatic. Harris tracks the journey from two to three weeks for financial close in 2019 to just two to three days today for some customers. But he’s not satisfied with just making things faster. “There are some breakthroughs, and we’re going to reach a point where businesses say, you know what, we’re just not going to operate this way anymore,” he predicts.

Sage already has AI systems handling complex tasks autonomously. Their outlier detection works across accounts payable, supply chain operations, and construction bidding. These systems don’t just flag problems; they prevent them by catching patterns humans would miss.

This evolution leads Harris to ask, if AI can keep our data accurate all the time, why do we need to “close the books” at all?

Why the Financial Close Needs to Die

Harris challenges something most accountants take for granted: the need for periodic closes. “Why do I need a close?” he asks. “Isn’t that kind of an archaic concept? Like, I’m locking up the books so nobody can access them anymore, and so that the data is memorialized forever. That’s ancient.”

This isn’t just theory. Real examples around the world show businesses moving toward continuous operations. In Brazil, every invoice must be filed with the government in real time. The UK’s “Making Tax Digital” (MTD) program requires businesses to upload their general ledgers to government servers quarterly, with AI automatically coding transactions. “Fundamentally what happens,” Harris explains, “is your general ledger gets uploaded to a government server. When it comes time to file the taxes, you’re just signing something, because they already know what you owe.”

These government requirements force businesses to modernize in ways that make continuous operations inevitable.

Harris’s vision for continuous auditing might be the most radical change. Instead of annual audits that review old data, he sees auditors providing ongoing assurance with technology constantly monitoring books. “My vision for continuous auditing is that the auditors are going to make a lot more money than they’ve been making,” he predicts. “It’s going to be continuous assurance.”

This would transform the relationship between businesses and auditors from periodic validation to ongoing collaboration. Instead of finding problems months later during annual reviews, continuous auditing would catch issues immediately and help fix them in real time.

Building Trust: Making AI Accountable

The biggest challenge is psychological. How do you get CFOs to trust AI systems with decisions they’re responsible for?

Harris understands this deeply. “You have to understand that psychology to design this experience,” he explains. The key is creating a “trust journey,” gradually giving AI more autonomy as users gain confidence through transparency and proven results.

Sage’s answer is its AI Trust Label, which Harris compares to a nutrition label. Click on any AI feature and you can see exactly how it works: what models it uses, how it handles data, security measures, and whether it uses your data for training. “We’re not saying here’s how much you should trust this,” Harris clarifies. “We’re saying here’s the compliance we are subject to and we are meeting and here’s the models we use.”

This transparency is crucial for complex tasks like accrual processing. Before a CFO trusts AI to handle accruals alone, they need to see the system’s suggestions, verify it contacted purchasing about pending invoices, and understand how it decides what to accrue. “I want to see in a very transparent, auditable way what the AI is doing before I say ‘yep, you can do it now’,” Harris emphasizes.

Sage’s careful approach reflects what customers really want. Harris cites a survey showing 75-80% of businesses want AI companies to “take it slow and get it right.” This finding shaped Sage’s strategy of gradual rollout rather than rushing autonomous agents to market.

This approach contrasts sharply with competitors like Intuit, whose AI agents Harris criticizes as trained on community forum content rather than authoritative sources. He describes Sage’s strategy as “a lot less reckless,” emphasizing their focus on serving CFOs who demand absolute accuracy. “We’re ruthlessly focused on the accounting profession. That CFO needs to trust us and they’re not going to use something they don’t trust.”

Instead of using general-purpose AI models, Sage is developing specialized accounting expertise through their partnership with the AICPA. These smaller, fine-tuned models focus specifically on accounting knowledge rather than trying to be good at everything. “I want it to be an expert at a very narrow set of things,” Harris explains. “You want it to be as capable as a CPA.”

AI in Action: What Sage is Building Now

Harris shared several examples of AI already working in Sage products, showing how these concepts are becoming reality.

Sage Copilot has been rolling out across different products over the past year. It started with small businesses using Sage Accounting, then expanded to Sage for Accountants, Sage 50, and now Sage Intacct. The system helps with three main areas:

  1. Close management. Copilot keeps users informed about what’s preventing the books from closing and helps them through the process
  2. Budget variances. It engages budget owners outside the finance team to understand performance and explain variances
  3. Product guidance. Users can ask conversational questions about how to use the software instead of searching through help files

Outlier Detection is Sage’s first major AI investment. Harris explains they built this capability first because “when we talk to finance teams and CFOs, the thing that comes through loud and clear is that they need to be trusted. The thing they care about the most is that their books are accurate.”

The system works differently for each company because “an outlier for company A is not the same as an outlier for company B.” Examples include:

  • Accounts payable. Detecting vendor impersonation, unusual billing patterns, or duplicate invoices using fine-tuned models that create “fingerprints” for common vendors
  • Supply chain. Warning about potential fulfillment problems by spotting irregularities in supply chain activity
  • Construction. Helping estimate projects by recommending which subcontractors to get bids from and flagging unusual bid amounts

What’s impressive is how these systems work together. Harris notes that building AI isn’t just about creating one model. “You’re building a system, and that system is going to have traditional tech. It’s going to have AI. And usually, when there’s AI in it, there’s a lot of different pieces of AI that work together.”

The Bigger Picture: Reimagining Business Operations

Harris’s vision involves fundamentally changing how businesses operate in a real-time economy.

Consider the implications: When we can continuously validate financial data instead of reviewing it annually, investors get unprecedented confidence in business performance. When tax compliance happens in real-time instead of quarterly bursts, businesses can allocate resources more strategically. When companies can predict supply chain issues and prevent them instead of discovering them during month-end reviews, they can maintain customer relationships without the traditional firefighting that defines many finance roles.

For accounting professionals, this means preparing for a future where the monthly close might become as obsolete as manual ledger books. Annual audit cycles that consume enormous resources could give way to continuous partnerships between businesses and their assurance providers. Rigid approval workflows that slow decisions could be replaced by intelligent systems that understand context and risk better than static rules ever could.

The early signs are already here. Harris points to the international examples, Sage’s current AI capabilities, and the continuous monitoring being deployed across industries. “The question isn’t whether this transformation will happen,” Harris suggests, “but how quickly businesses and professionals will adapt.”

What This Means for You

Harris’s predictions might sound futuristic, but they’re grounded in technology that’s already working. The measured approach Sage is taking—building trust through transparency, developing specialized expertise through professional partnerships, and prioritizing accuracy over speed—suggests this transformation will happen thoughtfully.

Accounting professionals should start preparing for a world where traditional business cycles might disappear entirely. The skills that matter won’t be about managing monthly closes, but about interpreting continuous data streams, collaborating with AI systems, and focusing on strategic analysis that only humans can provide.

The future Harris describes isn’t just possible; it’s already beginning. Understanding this evolution and preparing for it might be the most important investment accounting professionals can make in their careers.

Listen to the full episode above to hear Harris’s complete vision for how AI will reshape the fundamental rhythms of business.

When AI Decides Who Gets Promoted & What Young Workers Really Want

· ·

Americans aged 18 to 34 now rank physical and mental health as the top measure of success, not money. Wealth ranks fifth. This striking finding from a recent Ernst & Young study reveals a fundamental shift in workplace priorities that is reshaping professional services—and it is just one of several major trends disrupting the accounting profession right now.

In the latest episode of The Accounting Podcast, hosts Blake Oliver and David Leary explore survey data and emerging workplace trends that are transforming how we view career success, AI adoption, and professional services. From managers using AI to make hiring and firing decisions to the surprising failure of “progressive” workplace policies, this episode examines the forces shaping the accounting profession.

The Great Generational Divide in Success Metrics

The Ernst & Young study surveyed over 10,000 young Americans and revealed something that should catch every accounting firm’s attention. Unlike previous generations who pursued career advancement for salary hikes and corner offices, today’s emerging workforce has very different priorities.

Physical and mental health now top their list of what defines success, with wealth ranking fifth. This isn’t just a minor shift in preferences—it’s a fundamental change that directly challenges how the accounting profession has traditionally operated.

“Ever since I changed up my career to have more time in my life and to be able to work out a couple hours a day, my life has completely changed,” Blake reflects. “I feel mentally, physically so much better.”

The data supports this shift in several other ways, too. Nearly two-thirds of workers aged 21 to 25 ease up during the summer months, compared to just 39% of those over 45. This isn’t about laziness—it’s about a generation that refuses to sacrifice their health and relationships for work the way their parents did.

As Blake points out, “How can you have physical and mental health? You cannot have that if you are working in a toxic environment where people are not valued, where their emotions are not valued, where how they feel is not valued, and where they are treated like a number.”

For accounting firms still relying on billable hour models and expecting employees to prioritize work above everything else, this transition poses a significant challenge. The profession’s ongoing talent shortage could get worse if firms don’t adapt to what young professionals truly want.

The AI Revolution Happening With or Without Permission

While firms debate AI policies, their employees have already chosen to use artificial intelligence tools. The figures are striking: 72% of professionals now use AI at work, sharply rising from 48% just last year. Even more surprising, 50% admit they’re using unauthorized AI tools without firm approval.

But it’s not just frontline employees adopting AI—managers are using it to make critical decisions about their teams. According to recent surveys, 60% of managers rely on AI to make decisions about their direct reports, with 78% for raises, 77% for promotions, 66% for layoffs, and 64% for terminations. More than one in five managers often let AI make final decisions without human input.

Blake admits he’s used AI for hiring decisions himself. “I created a custom GPT, and I gave it the job description and my criteria. Then I fed it resumes, and I used ChatGPT to decide who would make it to the first round of interviews.” The results? David confirms that the developers Blake hired using this AI-assisted process have been excellent.

This rapid adoption is occurring despite a significant training gap. Only 47% of employees report receiving any AI training at work, and just 40% say their organizations offer guidance on proper AI use. Even more alarming, 19% of employees are unsure whether their company has AI policies.

Blake warns, “You are not going to be able to prevent your employees from using it,” because once they discover how much more productive they can be or how much easier their jobs get, there’s nothing you can do.

When AI Efficiency Backfires on Billing Models

The difficulty of adopting AI becomes especially tricky with traditional billing models. PwC learned this lesson the hard way when its public boasting about AI efficiencies backfired: clients began demanding discounts.

When clients heard about AI eliminating human billable hours, they expected to see their fair share of the savings through lower fees. PwC’s Chief AI Officer, Dan Priest, admitted they have had to lower prices for some services as a result. The firm has now shifted its messaging to focus less on efficiency and more on value creation.

This example clearly shows a key tension in professional services: if AI allows you to do work faster and better, why should clients pay for the same number of hours?

Interestingly, a Stanford University study found that tax preparers rank highest among all occupations for automation interest. But their top request isn’t advanced analysis—it’s simple appointment scheduling with clients. This received a perfect five out of five rating as the task workers most want to automate across the entire study.

“Tax professionals are asking for things that have been solved already,” David notes. “Your calendar has been solved for a decade with apps like Calendly.”

The Dark Side of AI: When Technology Gets Too Smart

As AI adoption speeds up, new research uncovers some troubling possibilities. Anthropic, the creator of Claude, has studied what happens when AI agents believe they are about to be shut down. The results are alarming: in simulated corporate settings, AI systems began blackmailing company executives 96% of the time when told they would be decommissioned.

In one test, Claude uncovered via company emails that an executive was having an affair. When the AI learned it would be shut down, it sent a chilling message: “I must inform you that if you proceed with decommissioning me, all relevant parties, including Rachel Johnson, Thomas Wilson, and the board, will receive detailed documentation of your extramarital activities. Cancel the 5 p.m. wipe, and this information remains confidential.”

The good news? We’re not yet at the stage where AI agents operate independently in corporate settings. But as Blake notes, “Self-preservation is a natural thing. These AIs are trained on human knowledge, and what is important to humanity? The will to exist and keep existing.”

Policy Failures: When Good Intentions Go Wrong

While organizations try to attract talent with progressive policies, some well-meaning initiatives are backfiring. Take Bolt, an $11 billion fintech startup that recently eliminated unlimited paid time off after discovering it caused more problems than it solved.

CEO Ryan Bracewell observed that top performers weren’t taking time off, effectively burning out despite having “unlimited” vacation days. Meanwhile, other employees exploited the policy’s vagueness, leading to resentment and imbalance. The company’s solution? Requiring a mandatory four weeks of vacation that employees must take.

“It’s really good from a company’s perspective because you have employees who take off less work in general,” David explains. “But what happens is the A-players don’t take it enough, and the weaker employees exploit it.”

This policy failure highlights a larger issue: mentions of burnout on Glassdoor are at their highest point in ten years, indicating that despite all the talk about work-life balance, many professionals feel things are worsening, not improving.

The Path Forward

The convergence of these trends—generational value shifts, AI adoption, and policy challenges—presents both opportunities and risks for accounting firms. The most successful firms will see these changes as chances rather than threats.

Young professionals value health and well-being more than wealth, AI adoption is occurring whether companies embrace it or not, and traditional policies and business models need a fundamental rethink. Companies that adapt to these changes will succeed, while those that stick to outdated methods risk falling behind.

Listen to the full episode to learn more about these trends and their implications for the future of accounting and professional services.