ACC: 74% of US workers worry more about AI dependency

The Acceleration Community of Companies (ACC) published national research on Aug. 19, 2026, finding that 74% of U.S. workers are more concerned about becoming overly dependent on AI tools than about being replaced by them. The figure, ACC says, highlights a shifting set of employee anxieties as employers roll out generative and assistive AI across routine and knowledge work.

ACC described the survey as a national snapshot of worker sentiment about AI in the workplace and framed the dependency concern as distinct from fears about job displacement. Respondents, the organisation said, signalled that their top worry is over-relying on tools for decision-making, tasks and professional judgment — a problem that HR leaders must weigh against productivity gains when deploying AI-assisted systems.

For human resources executives, the finding reframes common adoption debates. Rather than focusing solely on headcount impact, people teams now face questions about training design, monitoring practices and safeguards to prevent skill erosion. ACC said employers should consider structured reskilling, clearer policies on when and how employees should use AI, and metrics to ensure staff retain critical capabilities while using automation.

The survey also touches on privacy and monitoring tensions. Workers who fear dependency may resist audits or surveillance intended to measure reliance on tools, creating trade-offs for HR between transparency and oversight. ACC noted that governance frameworks — including role-based controls, logging of AI suggestions and human-in-the-loop checkpoints — can be used to balance autonomy and accountability as firms scale AI across functions.

The report arrives amid rising regulatory attention to workplace AI. Regulators in the U.S. and the EU are increasingly focused on how automated systems affect fairness, transparency and worker rights; the EU’s regulatory framework on artificial intelligence is poised to impose obligations on higher-risk uses of AI. That policy backdrop adds urgency for employers to document how they train, validate and monitor AI systems used for hiring, performance management and operational decision-making.

ACC’s data point also echoes broader questions about the design of assistive systems. Human-in-the-loop configurations mitigate some dependency risk but require cultural changes — managers must calibrate expectations about when employees should accept, override or ignore AI suggestions. For unions and employee representatives, perceived overdependence could become a bargaining point as workplaces formalise AI governance.

What ACC did not disclose in the public materials were methodological details that HR teams typically seek: the sample size, sampling frame, weighting, demographic or industry breakdowns, and the precise wording used to define "overdependence." The group also did not publish raw data or independent audit results that would allow third parties to validate subgroup patterns, nor did it outline tested interventions employers could adopt to reduce dependency risk.

The attention to dependency rather than displacement signals a subtle but important shift for people leaders. As organisations expand use of generative and assistive models, HR will increasingly be tasked with reconciling productivity targets and worker competence: designing training that preserves core skills, setting governance that limits uncritical reliance on machine outputs, and creating privacy-respecting monitoring that employees will accept. How employers respond over the next 12–24 months will shape whether AI is integrated as a trustworthy assistant or becomes a source of deskilled, brittle workflows.