AI Adoption Erodes Employee Judgment, BCG Survey Finds

Companies track AI usage but miss cognitive costs

Organizations rolling out artificial intelligence tools obsessively measure adoption rates—how many employees log in, how frequently they use the systems, how much time they save. What almost no one tracks is what those tools are doing to workers' ability to think critically.

A Boston Consulting Group survey of 70 C-suite executives, first reported by Quartz, reveals that half already see judgment and problem-solving skills deteriorating within their organizations. More than 60% believe this erosion will damage business performance within three to five years.

The skills at risk—judgment, problem framing, creative thinking—are precisely the capabilities that drive long-term competitive advantage. They're also the ones leaders say AI threatens most directly.

Workers stop verifying AI output

Nearly 90% of surveyed executives told BCG that employees have stopped checking AI's work. Polished drafts and confident analyses arrive without anyone questioning whether the underlying reasoning holds up.

When verification stops, accountability follows. More than half of leaders report that employees are less willing to own their decisions than before AI implementation. Nearly half observe more groupthink, as teams run identical problems through identical tools and accept identical answers. Forty-three percent say their people argue less about whether those answers are correct—meaning mistakes go uncaught.

A separate BCG study of AI training programs found that completion rates tell companies little about actual skill development. Employees pass assessments and demonstrate understanding in training, then return to their desks and either revert to old habits or lean harder on AI to do their thinking.

Entry-level learning paths disappear

New hires face a particularly acute problem. AI now handles the research, drafting, debugging, and problem-solving tasks that used to train junior employees. When leaders decide which tasks to automate, they're effectively deciding which skills the next generation will never build.

With simpler analytical work automated, new hires are expected to perform work that previously required five to 10 years of experience. Companies have eliminated the entry-level assignments that once taught people how to do the job, without replacing that learning pathway.

Senior employees aren't immune. When experienced professionals get busy, they start letting AI make calls they used to make themselves. One leader described this as the "autopilot" pattern—work still looks sharp, so no one notices fewer real decisions are being made.

A third of surveyed leaders said senior employees are teaching less than they used to. People go straight to language models for answers instead of asking colleagues who already have context. The explanatory conversations that used to convey not just what to do but why happen less frequently.

Brain scans reveal cognitive decline

The cognitive impact is measurable. A 2025 MIT study found that participants who used ChatGPT to write showed significantly lower brain activity than those who used web search or worked independently. They also remembered less of what they'd written afterward.

Why it matters

As AI tools become cheaper and more capable, the temptation to automate more cognitive work will intensify. Organizations that fail to preserve employees' thinking skills risk creating workforces that can execute but not strategize, that can polish but not originate. The companies that maintain competitive advantage will be those whose employees can still think without machine assistance—a capability that requires deliberate protection, not just adoption metrics.

A handful of companies in BCG's research have begun redesigning workflows to preserve cognitive skills. At Shell, junior employees work through problems independently before using any AI tool—framing questions, testing assumptions, and producing initial analyses by hand before using AI to refine their work. Early results show these employees explain their reasoning more clearly and require less managerial support.

The findings were first reported by Quartz, based on Boston Consulting Group's survey of C-suite leaders and separate research on AI training programs.