Karbon's Lead Product Marketing Manager, Lexi Beausoleil, calls this the AI slingshot moment. A slingshot only works if you're willing to pull back first.

The rebuilding phase requires discomfort. You need to shine a light on everything that's broken and ask hard questions about where work is being done manually that doesn't need to be, where workarounds have been built on top of workarounds, and where institutional habit has been masquerading as process. 

That kind of audit is slow, and most firms avoid it because the day-to-day doesn't stop while you're doing it.

But the firms willing to do it are loading the slingshot.

The accounting industry has been here before. Think about practice management software adoption. The firms that invested in genuinely building those systems—migrating data, retraining staff, rethinking how work moved through the practice—had to slow down to do it. First it was extra work. Then it made sense. And now there is a measurable and still-widening gap between the firms that did that work and the firms that didn't.

The same lesson is playing out again, at a much larger scale, with AI.

What this means in practice: if your team is feeling stretched right now, it isn't necessarily a sign that AI isn't working. For many firms, it's a sign that the real work has started. The exhaustion that comes with genuinely rebuilding is different from the exhaustion of doing the same things the same way. One leads somewhere. You need to understand which kind you're in.

Training, strategy, and policy reduce exhaustion

Only 46% of firms are actively investing in AI training for their teams, and just 21% have a documented AI policy or AI strategy. That gap is where a lot of the well-being opportunity is hiding.

Training changes the numbers in serious ways. Advanced AI users save 82 minutes a day, compared to 48 minutes for beginners. That's a 34-minute gap between someone who has been trained and someone who hasn't.

Governance compounds those gains further. Firms with a documented AI policy save 17% more time than those without one, and firms with an AI strategy save 20% more. You might think that a strategy and policy slows AI adoption. But in reality, it’s giving team members the guidance they need to explore it.

Without this, they’re left learning by trial and error. And a lot of the time, they can suffer decision paralysis and overwhelm. What’s the best way to use AI? What’s going to make the biggest impact? Which tool is the best?

Recommended reading: How to get started with generative AI at your accounting firm

The gap between leaders and their teams

AI sentiment is not one-to-one across roles in an accounting firm. Firm leaders (partners, directors, owners) are consistently the most optimistic. They tend to use AI for communication, strategic planning, and high-level ideation, where it performs well and the wins are easy to see.

Individual contributors (accountants, bookkeepers, enrolled agents, etc.) have seen the largest increase in skepticism of any group. They're the ones using AI directly in daily workflows, like drafting SOPs, preparing tax work, and producing technical outputs. 

That hands-on exposure means they're also the first to encounter the parts about AI that still need work. The outputs that require careful review, the hallucinations, the extra validation steps before anything can go out the door.

And this is why they need to be included in conversations about how AI gets deployed, and why they’ll benefit the most from clear policies, ongoing training, and a genuine say in how the tools evolve within the firm.

A leader working in isolation about a firm’s AI approach and deployment is a risk. The report states that 33% of individual contributors and 37% of operations, technology and administrative staff are concerned about the future of their role as AI matures. That's a healthy professional conversation to be having, and leadership teams that engage with it openly, rather than avoiding it, tend to build more confident and capable teams.

What’s on the other side?

The well-being gains from AI are real. They're just not automatic, and for firms in the middle of the slingshot pullback, they're not immediate either.

But here's what's waiting on the other side. Think beyond faster workflows.

Firms that have genuinely rebuilt their operations around AI are running more advisory capacity, communicating more proactively with clients, and offering their people something that has been structurally absent from the accounting profession for decades: a different deal.

The implicit bargain that a career in accounting has historically required—your nights, your weekends, your family dinners, your health, entire seasons every year—is one the next generation has largely declined. The firms that rebuild now are doing so to offer their people a career worth having.

As Mike Libbey, Partner & COO at YBL and 2025 Karbon Excellence Award Winner for AI Innovation, puts it:

“The firms that lean into the change brought about by AI will move faster, operate with more clarity, and attract the next generation of talent. The firms that resist it will feel like they are swimming upstream.”