Are Workers Fighting AI — Or Struggling to Keep Up
NOTE: We are seeing wildly varying numbers on adoption and resistance, so we advise that you analyze the data and not just assume it is accurate. Part of this is based on whom they are interviewing – management vs. individual contributors.
Despite widespread anxiety about artificial intelligence displacing jobs, workers aren’t actually pushing back on the technology. Only 2% of tech leaders report significant employee resistance to AI agents, according to new data from KPMG — a striking figure given how loudly the debate rages in the broader culture. Employee adoption of AI agents has already hit 68%.
So if workers aren’t resisting, what’s the real problem? Largely, it’s a skills gap. Among leaders who did report pushback, 78% traced it to a lack of training and fears about job security. More than half also pointed to concerns around trust and safety, with some workers reporting that AI is actually increasing their workloads rather than reducing them.
These anxieties aren’t irrational. Employers are increasingly citing AI in layoff announcements, and entry-level hiring continues to contract. Workers have every reason to wonder what this technology means for their careers — even as they adopt it.
From “Will They Use It?” to “They Have To”
KPMG’s findings suggest organizations have largely moved past the question of whether employees will embrace AI. The workplace is now entering a new phase where adoption is simply expected. Nearly half of tech leaders say AI literacy has become a workforce priority, with many companies introducing mandatory training, usage requirements, and performance metrics tied to AI adoption.
As KPMG’s US advisory leader Kevin Bogle put it, companies need to rethink how work gets done by embedding human-machine collaboration into everyday tasks — not treating AI as an optional add-on.
The Gap Between Vision and Reality
That said, the industry’s grand vision of autonomous AI “coworkers” is still far from today’s reality. A separate KPMG study found that 63% of organizations require human review of AI agent outputs. Current deployments are largely limited to low-stakes tasks: triaging IT support tickets, answering HR questions, handling routine workflows.
Workers may be comfortable with AI agents today precisely because those agents still function more like assistants than replacements. Whether that comfort holds as the tools grow more capable — and more autonomous — remains an open question.
On the business side, scaling AI comes with its own headaches. Organizations are deploying the technology without full visibility into costs, with average AI investments projected to reach $269 million over the next 12 months. Data privacy concerns and a shift toward lower-cost, higher-performing models are emerging as the dominant strategic priorities for the months ahead.
The bottom line: AI adoption in the workplace is happening faster and more smoothly than the headlines suggest. The harder challenge isn’t getting workers to use the technology — it’s making sure they have the knowledge, support, and security to use it well.
by 