Recently, one of my kids was frustrated with math homework. “Why do I have to memorize multiplication tables,” they asked, “if I can just use a calculator?”

It’s a fair question. After all, calculators are faster. And with smartphones, they’re almost always accessible. But as I explained, if you cut corners at the beginning, you never really learn the basics. You don’t build the foundation you’ll need as the subject becomes more advanced. You can’t master calculus unless you understand why 3 x 8 = 24.

That conversation got me thinking about how we’re approaching AI in the workplace. If employees jump straight to AI, they risk what experts call “ accomplishment hallucination .” It’s the latest workplace buzzword—the illusory feeling of achievement that comes from speeding through a task while skipping the hard thinking a problem requires.

Increased output can feel like increased productivity. But you may be selling yourself short. Without the fundamentals, you’ll never reach true mastery. It’s calculator math without number sense.

Today’s leaders can’t afford to ignore the risk of accomplishment hallucination. Here’s how to address it in your company.

Always Start With Independent Reasoning

If you’ve ever intended to use ChatGPT for a simple request and ended up delegating to the LLM way more of a task than originally planned, then you understand why AI use is a slippery slope. Most LLMs are designed to make you want to continue using them. You ask ChatGPT to generate marketing campaign headlines; it churns them out in 10 seconds, and then asks, innocently enough: Would you like me to outline a marketing email using one of the headlines? Pretty soon, you’ll have delegated the entire task to AI.

At Jotform, we have a hard line about AI use. Before using AI, employees have to articulate their own ideas. Take a stab at it by themselves and use AI as a collaborative tool—like an editor or colleague. Humans are the leaders, AI tools are the helpers—not the other way around.

For example, an employee can draft a hypothesis and then ask an AI tool for feedback. Or, they can outline a solution before prompting AI to expand it to other use cases.

It’s a simple but powerful rule that protects the mental work that builds expertise.

Institute AI-Blackout Windows

There are few things more frustrating nowadays than asking a colleague for input and receiving what is clearly an AI-generated reply. If I wanted ChatGPT’s insight, I could ask on my own.

But the hard truth is that in the past couple of years, it’s become commonplace to consult AI tools as a knee-jerk reaction to answer questions and find solutions. The responsibility lies with leaders to institute designated moments during which employees completely refrain from AI use—I call them “AI-blackout windows.”

For example, during the first half of any strategy season, we institute an AI-blackout window. During our weekly demo days, too. These are the times when we hash out ideas and look for solutions together. The group dynamic helps us to build creative momentum. These sessions have led to some of our most successful product innovations. Had we just outsourced the work to an AI tool, I doubt we’d have these kinds of breakthroughs.

Going through the motions of thinking through ideas, looking at different perspectives, and playing devil’s advocate isn’t a waste of time—it continues to build creative muscles.

Cold-Call Explanations Sometimes

In college, I dreaded being cold-called during huge lectures to explain my reasoning behind an answer. But looking back, I understand the value. Walking step-by-step through my process not only revealed that I did the work myself—it wasn’t just a test of the honor system. It also helped me to build confidence in my answers.

At Jotform, we encourage employees to leverage AI tools. But they also know that they may be “cold-called” upon to reason through AI-generated outputs. They should be able to explain why an answer makes sense and identify any possible assumptions. We might ask about potential weaknesses and if conditions change, how the answer might shift as well.

It’s not a punishment, and it’s not a lack of trust in our team members. The point is to ensure they’re confident in all of their decisions, including when AI helps to inform them.

Safeguard The Apprenticeship Layer

Junior employees today, especially from Generation Z, find themselves in a unique position. Employers are eager to capitalize on their digital native skills; their easy fluency with emerging tech and AI. But the very capabilities that make them valuable can also hinder their learning, especially the foundational skills. One Reddit user went so far as to call AI a “ death trap ” for junior developers. I understand the genuine concern.

Leaders can help curb the risk of junior employees skipping the fundamentals by pairing AI use with mentorship. AI use should never take place in the dark, full stop. Instead, organizations can assign mentors to junior employees to chat about how they’re using AI, give them tips for using it effectively, and help them choose tasks to complete without it—to get a true gauge of their progress. For example, advise junior developers to generate code themselves and use AI to refine and troubleshoot. By developing it themselves, they better understand how to fix it, if it fails, and anticipate possible improvements in the future.

Importantly, the goal isn’t to put junior employees at a disadvantage. Rather, it’s to ensure they’re building their skills and becoming better at their jobs, not merely faster.

Reward Output Quality, Not Volume

This final strategy is a more subtle shift. It’s built into the fabric of an organization, not instituted through a policy or technique. And it starts at the top: cultivate a culture where quality is prioritized over quantity of output.

You might be wondering: How do I do this in real life? It’s all about the everyday practices that accumulate to signal what your company values.

When evaluating performance, consistent accuracy should be rewarded over mere speed. Employees should be recognized for meaningfully showing up—being good colleagues, bringing positive energy and ideas to work—not for clocking the longest hours. Being productive is important, but being productive and clearly explaining your insights to the team is even more important—because when one person improves, the whole team can rise with them.