What if Your Team Doesn’t Trust AI?

Why addressing skepticism is part of building better processes

AI presents different problems depending on where you’re at in your accounting career.

If you’re new to the profession, you might not have enough experience to recognize when AI gets something wrong. You can question the output, but how do you know what to question when you’re still learning the fundamentals?

If you’ve been an accountant for 30 years, you might have the opposite problem. You’ve spent your entire career developing expertise and learning to trust your own judgment. Now, you’re being asked to rely on technology that can produce an answer in seconds, and you’re not convinced it knows what it’s talking about.

With so much pressure to adopt AI and become more efficient, I think we’re overlooking how stressful that can be for people at every experience level.

Skepticism Can Show You What Needs Fixing

We need to make room for those concerns instead of rushing everyone through the learning process. Ignoring people’s concerns won’t make them disappear. And more importantly, working through them can help firms identify where their processes need to improve.

If younger accountants are worried about recognizing mistakes, what training and review procedures do they need to feel confident? If experienced accountants don’t trust the output, how can we give them opportunities to test it and understand where it performs well?

These are the conversations firms need to be having as they incorporate AI into their workflows. Sometimes these conversations can get pretty uncomfortable, especially when people have different levels of experience and confidence with technology, but they’re how we build processes that people understand and trust.

I want accountants to feel comfortable questioning AI. I also want them to feel comfortable questioning the way their firms are using it.

Give Your Team a Way to Build Confidence

These three practices can help your team work through skepticism and build confidence in how your firm uses AI.

1. Review AI-generated work together. Have newer and experienced accountants evaluate outputs side by side. Compare the results against source material, question the assumptions and discuss how you know whether an answer is reliable.

2. Test AI on familiar work. Use completed assignments with known outcomes to evaluate AI’s performance. Document where it succeeds, where it makes mistakes and what needs human review before the work can be trusted.

3. Make concerns part of the conversation. Ask your team what makes them uncomfortable about using AI. Discuss mistakes, unexpected results and questions about reliability. Use what you learn to improve your training, review procedures and expectations.

The goal is to make questioning, testing and learning part of how the work gets done, so your team has a reason to trust the process.

Partners Have to Do the Work Too

If your team doesn’t trust AI, get curious about why. They may be pointing out a gap in your training, a weakness in your review process or an expectation you haven’t clearly explained. That’s information you need to build a better firm.

But skepticism can’t be where the conversation ends, either. “I don’t trust it” should lead to testing, questioning and learning. That applies to the newest staff member and the partner who’s been doing this for 30 years.

You don’t build an AI-first firm by telling people to trust the technology. You build it by giving them a way to develop their judgment and expecting everyone, including the partners, to do the work.