Nearly nine in 10 U.S. office workers now use AI at work, but only one in four have fundamentally changed how they work, Fyxer disclosed in its June 23 report based on a May 2026 OnePoll survey of 2,000 U.S. office workers.
“The AI adoption race is over,” Fyxer CEO Rich Hollingsworth said in the announcement. “Almost everyone is now using AI. Today’s challenge is helping people use it in a way that actually saves them time, significantly boosts their productivity, and genuinely transforms the way they work.”
The productivity paradox
The findings are complex. The same report found 69% of workers said AI had made them more productive, while 42% said it had increased their workload.
Fyxer presents that split as part of the same productivity trap: workers are adopting AI, but work is not always being redesigned around it.
Assessing the cost of the adoption gap
The gap between use and redesign is the core finding in Fyxer’s AI Productivity Trap report. Fyxer estimates that the gap could represent as much as $2.6 trillion in unrealized annual productivity across the U.S. workforce, but the figure is a modeled estimate rather than a measured gain already available to employers.
The company’s calculation uses 93 million U.S. office workers, an 88% AI-use rate, a 75% share of AI users outside its “AI Superworker” category and an average office-worker salary of about $68,766. Fyxer said the salary assumption is based on BLS data and described the estimate as indicative of productivity potentially left unrealized under current AI rollouts.
Integrated workflows vs. standalone tools
Fyxer draws its main distinction between integrated and standalone AI use, rather than between AI users and non-users. The report argues that workers using AI inside existing workflows report stronger results than workers relying on separate tools. It said about 83% of workers using integrated tools reported increased productivity, compared with 20% of workers using standalone tools.
The report calls the highest-performing group “AI Superworkers.” Fyxer said they make up 25% of AI users, with 87% reporting higher productivity. “The biggest gains come when AI becomes part of the workflow rather than another destination employees have to visit,” Hollingsworth said.
Email as a benchmark for automation
Fyxer applies the argument most directly to email. Its January Admin Burden Index said routine admin drains 5.6 hours per employee each week, with email topping the list of time-wasters. The June report says reading and writing emails remain the most time-consuming activities for U.S. office workers, while only 30% use AI to write emails and 22% use it to read them.
Fyxer sells an AI email assistant embedded into Gmail and Outlook, so its research supports the kind of workflow-level tool category it offers. The company’s survey still points to a broader employer problem: AI access alone does not show whether work has been redesigned, whether tasks have been removed or whether employees have simply added another tool to the day.
Separate workplace research, using different populations and definitions, points in the same direction. Gallup’s 2026 AI indicator found that 65% of U.S. employees at organizations using AI said it had improved productivity and efficiency, but only 12% strongly agreed that AI had transformed how work gets done in their organization.
McKinsey’s 2025 global AI survey also found that AI high performers were nearly three times as likely as others to have fundamentally redesigned individual workflows.
Management influence and worker autonomy
Fyxer’s report also raises a control problem. Workers who selected their own AI tools changed how they worked at nearly twice the rate of those using employer-chosen tools, 35% versus 18%, while 63% of managers and senior leaders chose their own AI tools compared with 42% of entry-level workers.