AI Productivity Paradox Traced to Leadership Failures, Not Tools

Employees using artificial intelligence regularly are saving the equivalent of an entire workday each week, yet aggregate productivity metrics show no corresponding gains. New research from Boston Consulting Group pinpoints the culprit: not the technology itself, but a fundamental breakdown in leadership communication.

BCG's 2026 Global AI at Work report surveyed nearly 12,000 frontline employees and found that 42% reported saving eight hours per week through regular AI use. However, 66% said they received limited to no guidance on what to do with that reclaimed time. Half are not redirecting those hours toward more strategic work.

Why it matters

The disconnect between individual time savings and enterprise-level productivity gains represents billions in unrealized value. Companies that fail to provide clear AI strategy and direction risk not only wasting investments in the technology but also fueling employee anxiety that inhibits knowledge sharing—the very behavior needed to scale AI benefits across organizations.

The tokenmaxxing trap

The research arrives as major technology companies confront the financial consequences of encouraging AI use without strategic guardrails. Amazon employees engaged in "tokenmaxxing"—maximizing their use of AI tokens to meet company metrics—a practice also adopted at Meta through AI user leaderboards. Microsoft canceled much of its Claude code licenses, while Uber exhausted its entire 2026 AI coding tools budget within four months.

"For my team, the cost of compute is far beyond the costs of the employees," Bryan Catanzaro, vice president of applied deep learning at Nvidia, told Axios. Microsoft AI chief Mustafa Suleyman acknowledged this week that Anthropic's pricing has become prohibitive.

David Martin, global leader of BCG's People & Organization practice, told Fortune that senior leaders are struggling to articulate clear vision and strategy around AI deployment. "This whole tokenmaxxing thing has probably run its course, and now it's hitting their cost base in a pretty big way," he said.

Amazon scrapped its internal AI usage tracking last week after employees deployed bots to complete meaningless tasks. "Please don't use AI just for the sake of using AI," Amazon senior vice president Dave Treadwell reportedly told staff.

Fear inhibits sharing

Beyond unclear direction, workplace anxiety around AI is preventing the knowledge sharing necessary for broad adoption. Recent BCG research found that when companies frame AI agents as digital employees rather than tools, it heightens displacement fears among workers.

"In a world where people are fearful, they are actually saying, 'Okay, how can I get a leg up on my peers, so that I can position myself to be highly valuable?'" Martin explained. "And because of that, they're a little bit reluctant to talk to their peers about what they're doing."

Companies that have integrated AI into their entire enterprise operating model while providing comprehensive upskilling training have seen greater success, Martin noted. Workers who feel empowered are more likely to share AI techniques with colleagues, accelerating organizational learning.

At the Fortune COO Summit this week, Okta COO Eric Kelleher said the problem reflects a failure of imagination. "Everyone has the mandate [to adopt AI]," he said, but managers remain focused on traditional metrics like headcount and org charts rather than reimagining work itself.

The findings were first reported by Fortune, drawing on BCG's 2026 Global AI at Work report and interviews with company executives.