Close Faster & Deliver 12% More Detail

AI is not replacing accountants, but it is enhancing their capacity to reduce repetitive work, enhance communication, and improve the accuracy of their reporting, according to a new study from Stanford University and MIT Sloan School of Management. It shows this technology can lift productivity and financial reporting quality when paired with professional judgment.

Widespread but Uneven Adoption

About 38% of accountants had integrated tools into their workflows, which is a notable increase from the 23% adoption rate reported just a year earlier, according to the study.

Usage is still uneven, however. The average accountant uses these tools a few times a week, while roughly 10% rely on them daily. It is often used for data entry, administrative work, and business communication, where automation can reduce manual effort.

Despite optimism surrounding automation systems, many accountants are cautious. Some respondents noted concerns about errors and data security, with 62% expressing worries about accuracy in reports. As one participant summarized in the study, “AI reduces our routine workload, but it also requires constant review and validation.”

Real-World Evidence

To take it a step further, the research team partnered with a San Francisco-based firm that builds and licenses accounting software. The firm’s technology integrates large language models with bookkeeping and reconciliation systems to automate transaction categorization and document processing. The goal is to complete these actions while keeping accountants in the loop.

These results support what many companies have long suspected: This technology is most effective when applied to repetitive, rule-based accounting tasks. Doing so allows professionals to focus on higher-level analysis and strategy. One of the quite tangible gains includes an 8.5% reduction in time spent on data entry (which freed about 3.5 hours in a 40-hour week).

Improvements in Financial Reporting Quality

Efficiency isn’t enough in accounting, though. Accuracy and compliance are still non-negotiable. The study found that AI adoption correlates with measurable improvements in the quality of financial reporting.

In fact, firms that used generative tools recorded:

• A 12% increase in general ledger granularity, and
• A 7.5-day reduction in monthly close time.

The combination of speed and precision suggests that it can enhance reporting processes rather than weaken them. “AI integration can enhance both the timeliness and precision of financial reporting,” the authors noted, adding that these benefits didn’t come at the expense of quality.

Complementing — Not Replacing — Human Expertise

A major insight from the research is that AI works best when paired with experienced accountants, those who can interpret and refine its output. The systems studied provide confidence scores that estimate how reliable each recommendation is. Experienced professionals tend to intervene when these scores are low, then use their own expertise to correct potential errors.

By contrast, less experienced accountants are more likely to accept suggestions without adjustment. This is why human oversight is a must.

The technology handles the routine work, while accountants ensure accuracy, context, and compliance. The balance of automation and expertise showcases what many firms are now calling the “human-in-the-loop” model.

Practical Steps for Your Team

Firms planning on intelligent software adoption can take several concrete steps to achieve similar goals to the ones documented in the study:

• Start small. Introduce tools for repetitive tasks such as data entry, expense classification, or reconciliations.
• Maintain oversight. Always review results and track confidence scores where available.
• Train staff. Offer short courses or internal sessions on interpreting outputs and managing exceptions.
• Measure outcomes. Track key metrics like close time, error rates, and time allocation across tasks before and after adoption.
• Protect data. Confirm that any platform complies with your firm’s privacy and security standards.

Implementing these steps helps ensure that automation provides value without introducing new risks.

What This Means for Finance Professionals

For accounting leaders, the takeaway should be relatively simple. Machine learning systems can deliver meaningful efficiency and reporting gains, but only when paired with human expertise and quality controls.

Instead of displacing accountants, this automation tool appears to amplify their effectiveness. It automates repetitive work, shortens reporting cycles, and enhances the accuracy of financial data. At the same time, it raises new expectations for professional judgment, review, and training.