Recent analysis from Glassdoor reveals a striking shift in how employees perceive artificial intelligence in the workplace. Between May 2025 and May 2026, mentions of AI in worker reviews surged by 240%, indicating that the technology has become a frequent topic of discussion. What is more telling is the direction of sentiment: while AI was praised in 81% of reviews back in 2019, only 43% of mentions were positive in 2026, with 53% now negative and a small remainder mixed. This reversal suggests that the initial enthusiasm surrounding AI is eroding as employees gain firsthand experience with its implementation across various industries.
The data show a clear correlation between job seniority and attitudes toward AI. Employees in director-level positions or higher expressed the most optimism, often viewing AI as a strategic asset that benefits the organization. Sales managers also reported favorable impressions, with roughly a two‑to‑one ratio of positive to negative comments. Product managers, recruiters, and software engineers likewise leaned toward approval, though their enthusiasm tapered as one moved down the organizational ladder. This pattern hints that access to decision‑making authority and the ability to shape how AI is deployed play a crucial role in fostering a positive outlook.
Why do leaders and certain specialist groups tend to see AI more favorably? For many, the technology augments tasks they already control—such as forecasting sales pipelines, screening candidate profiles, or generating code prototypes—thereby increasing efficiency without threatening their core responsibilities. Moreover, those in senior roles often feel insulated from the risks of displacement, surveillance, or increased workload that can accompany AI rollouts. Their positive feedback frequently centers on the organization’s ability to capitalize on the AI boom or to use the tools effectively internally, reflecting a perception of AI as a lever for competitive advantage rather than a source of personal risk.
In stark contrast, frontline and operational workers reported deep dissatisfaction. Insurance claims adjusters emerged as the most negative cohort, with nearly 98% of their AI‑related remarks expressing frustration. Similar levels of discontent appeared among accountants, customer service representatives, and IT specialists. These roles share common traits: they demand high accuracy, involve concrete problem‑solving, and place workers directly accountable for outcomes. When AI systems produce errors—misclassifying claims, generating faulty financial summaries, or misrouting service tickets—employees bear the brunt of correcting the mistakes, often while facing pressure from both management and clients.
The experience of insurance claims adjusters illustrates why AI can feel like a burden rather than a boon. In one documented case, an adjuster named Ahmad Jackson described how his employer’s AI‑driven loss‑reporting tool flooded his queue with misclassified claims, requiring manual rerouting. The system also produced hallucinated summaries that, when inadvertently shared with claimants, triggered complaints and eroded trust. Rather than reducing his workload, the technology added layers of verification and rework, ultimately prompting him to seek employment elsewhere. Such stories are not isolated; they reflect a broader pattern where poorly calibrated AI creates extra human labor to fix machine‑generated errors.
Beyond fears of job replacement—which still rank as the top concern but account for only about one‑fifth of negative remarks—workers cite several other grievances. Roughly 14% resent being compelled to adopt AI tools they had no say in selecting, while 13% believe their employers are prioritizing AI hype over core business objectives. Additional complaints include the use of AI for internal surveillance or communications (10%), perceived slowness in realizing benefits (9%), unrealistic expectations about ROI (8%), and anxiety about working in industries likely to be disrupted (7%). These findings suggest that the backlash is multifaceted, stemming from procedural, ethical, and strategic mismatches rather than a singular fear of automation.
The positive side of the narrative, meanwhile, highlights two main themes: organizational gain and internal efficacy. About 44% of favorable comments praised how the company was benefiting from the broader AI boom, such as through increased market valuation or new product offerings. Another 41% lauded instances where AI was applied well inside the firm—streamlining legitimate workflows, reducing mundane tasks, or enhancing decision‑making when appropriately governed. This split underscores that workers can appreciate AI when it delivers tangible advantages without compromising their autonomy or exposing them to undue risk.
Examining these trends through a lens of workplace power dynamics offers a compelling explanation. Employees who possess authority over when and how to use AI, and who feel secure enough that the technology will not jeopardize their livelihoods, tend to view it as an empowering tool. Conversely, those who lack agency and face direct accountability for AI‑driven outcomes experience the technology as an external imposition that amplifies stress and error‑correction labor. This divide mirrors broader socioeconomic patterns, where capital‑holding executives reap the perceived benefits of automation while labor bears the implementation costs and risks.
The workplace backlash against AI dovetails with larger societal movements resisting the unchecked spread of data‑intensive technologies. Protests against energy‑hungry data centers, community actions targeting surveillance‑laden wearable devices, and growing skepticism toward AI‑generated content all reflect a public unease with technologies that appear to serve corporate interests while diffusing costs onto individuals and communities. Within organizations, similar sentiments manifest as disengagement, increased burnout, and, in some cases, talent attrition as workers seek environments where technological change is negotiated rather than mandated.
For business leaders, the Glassdoor data serve as an early warning signal. Investing in AI without accompanying change management, employee training, and mechanisms for worker input can erode morale and trigger productivity losses that offset any efficiency gains. Companies should consider pilot programs that include feedback loops, allow opt‑out periods for poorly performing tools, and establish clear accountability frameworks that protect workers from being scapegoated for algorithmic mistakes. Transparent communication about the limits and purposes of AI can also help align expectations and reduce resentment.
Employees navigating AI‑heavy workplaces can adopt proactive strategies to safeguard their well‑being. Keeping a record of instances where AI generates errors or creates additional work provides concrete evidence for discussions with managers or HR. Seeking opportunities to up‑skill in areas that complement AI—such as data validation, AI ethics oversight, or process redesign—can shift one’s role from a victim of automation to a steward of it. Moreover, participating in internal forums or unions that debate technology policy helps collective bargaining power shape how AI is deployed.
Investors and market analysts should monitor employee sentiment as a leading indicator of long‑term AI ROI. High levels of worker dissatisfaction may presage higher turnover, increased training costs, and potential reputational damage, all of which can affect a company’s bottom line. Conversely, firms that demonstrate inclusive AI adoption—evidenced by positive Glassdoor trends, strong internal mobility programs, and measurable gains in both output and employee satisfaction—may represent more resilient investments in the AI era.