The Hidden Risk of AI: The Illusion of Expertise

The role of communications leaders has never been more strategic.

Today, we increasingly serve as strategic partners to our organizations, advising on issues that extend far beyond messaging. We help navigate ambiguity, anticipate risks, and make consequential decisions in an environment shaped by social complexity, geopolitical volatility, public policy shifts, evolving stakeholder expectations, and the accelerating disruption of AI.

We worked hard to earn our seat at the decision-making table. Yet as AI reshapes how work gets done and how expertise is developed, I find myself wondering: How will the next generation build the judgment and strategic perspective needed to keep that seat? Are we at risk of eliminating some of the very experiences that shape communicators into trusted advisors?

This concern crystallized for me after reading a recent MIT report on the impact of AI on education. One of its central warnings is that overreliance on AI is creating the illusion of learning.

The report sounds an alarm about “cognitive surrender” among students, as they increasingly offload thinking to chatbots at the first hint of struggle. The result is polished assignment submissions produced without the productive intellectual struggle of grappling with problems, developing arguments, or working through ambiguity. Yet it is precisely that “productive struggle” that builds genuine mastery, durable understanding, and intellectual maturity.

I believe the same underlying dynamic creates a similar risk in the workplace: the illusion of expertise.

As leaders, we are actively redesigning workflows around AI-enabled efficiencies, and the expectation to unlock productivity gains now reaches every level of the team. Communicators no longer need to start every project with a blank page and spend hours on research, building media lists, analyzing coverage, or synthesizing meeting notes. Much of this work can be outsourced to AI and completed in seconds.

These efficiencies are undeniably a tremendous benefit. 

The challenge is what happens when we remove the struggle without replacing the learning.

Some of the foundational activities we are automating have historically been how junior and mid-career communicators developed subject-matter expertise, pattern recognition, critical thinking, risk awareness, and contextual understanding necessary to grow into strategic counsel roles.

Now, the concern is that if we are not intentional about new pathways for development, we risk creating communicators who can produce polished deliverables and appear highly competent on the surface, yet lack the deep domain knowledge and battle-tested judgment required to navigate crises, respond to novel situations, and provide counsel to the C-suite when the answer isn’t obvious.

That’s the illusion of expertise.

Of course, the answer is not to resist AI or preserve inefficient work simply because that is how many of us learned. That’s not practical or possible.

The authors of the MIT report argue for the need to rethink how education is delivered and assessed in an AI-enabled environment. I see the same need in the workplace.

Here are some ways I am thinking about it in my approach to talent development:

• Create deliberate exposure to strategic decisions

AI can generate answers. It cannot replicate experience.

That makes it more important than ever to create deliberate opportunities for junior communicators to learn how organizations actually work. This means exposing them to business discussions, cross-functional decision-making, stakeholder tensions, executive debates, and moments of uncertainty.

The goal is not simply to have them in the room when decisions are made, but to help them understand the reasoning and trade-offs behind those decisions.

• Ask for thinking before deliverables

Before a team member starts working on an assignment and turns to a chatbot, find time to hear them articulate their own thinking first. Ask: What approach would you recommend we take? Why?

The goal is not to discourage AI use but to ensure people develop a point of view before an answer is generated for them. AI should strengthen thinking, not substitute for it.

• Make feedback about reasoning, not just editing

In a fast-paced environment, it's tempting to edit a draft and move on. But in an AI-enabled workplace, it is more important than ever to explore the thinking behind the work and explain the reasoning behind your edits.

Which stakeholders did you prioritize and why? What assumptions are you making? What alternatives did you consider? What trade-offs did you identify? 

Taking the time to debrief and ask these questions helps develop judgment over time, not just improve the immediate deliverable.

• Protect developmental work

Not every assignment should be optimized for efficiency. For activities such as narrative development, change communication, and crisis response, consider requiring a human first draft. 

The struggle of finding the right message and tone helps communicators develop the judgment to anticipate audience impact and reputational risk, capabilities that cannot be built through autocomplete.

• Expect teams to challenge AI, not just use it

One of the most important emerging skills is knowing how to challenge AI-generated outputs. Before presenting work for review, ask team members to identify flawed assumptions, missing context, reputational blind spots, and potential weaknesses in their recommendations. The goal is to build the habit of critical evaluation rather than accepting answers at face value.

• Measure development, not just productivity

Many organizations now track AI-driven efficiency gains. As leaders, we should also track learning.

In my team’s performance assessments, I pay attention to what new capabilities someone has developed, whether they are becoming more independent in their thinking, and what judgment they have demonstrated.

The bottom line for me is that preventing the illusion of expertise may require much greater intentionality in how we develop talent. We may need to bring a little more Socrates into how our teams operate: more questioning, more dialogue, and more critical self-examination of the work before it’s presented.

It may also require creating more opportunities to learn together. The MIT report flags that with increased reliance on AI, people may place less value on debating ideas and working through problems with others. We should consider the same dynamic in the workplace because expertise is developed not only through individual effort, but also through discussion and collaboration. Both can be even harder to cultivate in hybrid and remote environments.

AI can make us faster. It can help us produce more. Our job as today’s leaders is to make sure it doesn’t quietly cost the next generation of communicators the judgment and expertise they need to earn and keep their seat at the decision-making table as trusted advisors.