AI helps people get work done faster, but it can also lead employees to use some skills less often. Judgment, problem framing, and creativity risk eroding when too much of the thinking is left to the tool.
From individual habit to organizational risk
Tally sticks, calculators, and navigation systems are familiar examples of people handing the burden of thinking over to tools. With generative AI, this delegation covers an ever larger share of thinking. BCG calls this “distributed skill erosion.” When hundreds of people stop exercising the same skills at once, an individual habit becomes an organizational risk. This calls for a closer look at how work is organized around AI.
In a study of 70 senior executives, half already observe this erosion in their organizations. More than 60 percent think it will become a serious threat within the next three to five years.
The most critical skills are most at risk
The skills the research identifies as most important for long-term performance are also those most at risk: judgment and decision-making, understanding and framing problems, creative thinking, analysis and causal reasoning, and generating and evaluating solutions. When the ability to define a problem weakens, faster analysis cannot make up for it. Relational skills such as empathy, curiosity, resilience, and leadership appear more protected for now. They may also help preserve the skills at risk.
Signs of erosion within the organization
One early sign is accepting AI outputs without question. Approximately 90 percent of executives see this as the first warning sign. Ownership and accountability then weaken. “AI suggested it” becomes a way to evade responsibility for a decision. As ideas produced with the same tools grow more alike, debate and diversity of thought also diminish.
53 percent of executives say younger employees' development is slowing. When AI takes over the foundational tasks that build judgment, employees lose the repeated practice through which they gain experience.
Six ways to reduce the risk
According to the research, only one in ten companies has launched an organization-wide strategy or a targeted initiative on this issue. The six approaches highlighted in the research range from organizing work to designing the tools people use.
The first approach is to set conditions for use. This means clarifying which tools are approved, how outputs will be verified, and where human review is mandatory. It also includes defining AI-free zones for work that requires originality. The second approach focuses on how work is done. Tasks are divided between people and AI with care to preserve the master-apprentice relationship.
The third approach includes human skills in performance reviews. Both the result and how it was reached are assessed. The fourth adds opportunities to exercise skills to everyday work. AI-free problem-solving sessions and structured discussions are examples.
The fifth approach teaches employees to work with AI beyond a one-way question-and-answer exchange. The tool is used to generate better questions and challenge thinking. The sixth embeds thought-provoking prompts in the tools. Questions such as “What assumptions are you making?” invite users to think before receiving an answer.
The Kaykayoğlu perspective
The skills the research identifies as at risk overlap with areas we have focused on in our work for years. Problem framing, systematic creativity, and judgment in decision-making can be exercised through TRIZ and structured problem-solving methods. This is why we take human skills as the starting point for AI-Native transformation. AI can accelerate thinking, but when it replaces thinking, employees have fewer opportunities to use these skills. AI-free zones for tasks requiring originality and judgment, review steps that question outputs in important decisions, and regular problem-solving sessions are concrete ways to preserve those opportunities.
Sources
This text was prepared by reviewing the relevant sources and drawing on their findings.
