If AI Builds Your Spreadsheet, Who Checks It?

Many employees are quietly deciding they no longer need to improve their Excel skills. Their reasoning is simple: if Copilot, ChatGPT or Claude can write the formula, build the summary and draw the chart, why spend time learning how it works?

Arnold Muscat, Managing Director of College Africa Group (CAG), believes this is one of the most risky assumptions in the workplace today. "AI can produce a spreadsheet in seconds. What it cannot do is tell you whether the answer is right," he says. "If you don't understand the formula, you can't check it. And if you can't check it, you are trusting a number you don't understand."

Confident is not the same as correct

AI tools present their output with confidence. A formula looks professional, a pivot table looks complete and a summary reads well. But AI can misread a column, include the wrong range, ignore blank rows or apply logic that doesn't match how the business actually works.

"The danger isn't that AI makes mistakes. People make mistakes too," Muscat says. "The danger is that the person using it has no way of spotting the mistake. A wrong total in a management report looks exactly the same as a right one."

Core skills are the safety net

CAG teaches a simple framework for using AI with business data: Source Data → Process → Required Output → AI → Validate Against the Source. The final step is the one most often skipped, and it depends entirely on the user's own skills.

"To validate AI output, you need to know what a correct lookup looks like, how totals should reconcile and where errors usually hide," Muscat explains. "Those are Excel skills. AI hasn't made them less important. It has made them more important, because more output is being produced, faster, by people who didn't build it themselves."

What managers should ask

Muscat suggests managers ask a few practical questions about how their teams use AI with spreadsheets:

- Can the person explain the formula AI gave them?

- Have they checked the result against the source data?

- Would they notice if a range or condition was wrong?

- Is anyone reviewing AI-built reports before they reach management?

"If the answer to these is no, the problem isn't the AI tool. It's the skills gap underneath it," he says.

Use AI, but keep the skills

Muscat is clear that the answer is not to avoid AI. "Use it. It saves real time. But treat it like a capable new colleague whose work still needs checking, and make sure your people have the skills to do that checking."

CAG's Advanced Excel training helps finance, reporting and operations teams build the formula, data and reporting skills needed to work confidently with large workbooks and to check the results AI produces. Courses are hands-on and run on request for teams. 

Review AI platforms https://collegeafricagroup.com/ai-and-chatgpt-for-business-south-africa/

About College Africa Group

College Africa Group (Pty) Ltd is a South African corporate training, consulting and business productivity company established in 2003. CAG works with corporates, HR, finance, project and leadership teams across South Africa and Southern Africa, specialising in Microsoft Excel, Microsoft 365, Microsoft Copilot, AI in the Workplace and business process improvement.

Contact: Arnold Muscat, This email address is being protected from spambots. You need JavaScript enabled to view it.