Navigating AI in the Workplace: 2026

Among the workers who identified the use of these adoption strategies within their organizations, the most effective methods were offering monetary incentives (64%), providing multiple training sessions (63%), and organizing competitions (62%) to drive AI engagement. Monetary incentives and competitions are less commonly implemented compared to other strategies, but their impact is significant. Workers appear to better understand and embrace AI in collaborative environments where clear objectives are set and skills development is prioritized. A clear advantage of the ”carrot” approach (as opposed to the “stick”) appears to be in practice with the effectiveness of these programs, because reward-based strategies were seen as far more impactful than negative ones.

Integrating AI into existing — or redesigned — workflows is always going to present a challenge, and it brings an almost-unanswerable question: “How much AI assistance should I use?” Looking at median responses, workers provided directional guidance: 43% is optimal, while the acceptable range is between 22% and 59%. We refer to this range as the “Goldilocks Zone.”

At the heart of AI — and why so many leaders seek to implement it — is the promise of increased productivity. Productivity itself is difficult to self-quantify, but we can better understand it by examining the factors that shape the feeling of being productive in day-to-day work.

Self-reported assessments of AI experience help us understand the landscape of AI users in the workforce, but SHRM research sought to gain a better understanding of workers’ AI knowledge. As part of a forthcoming report, we developed the SHRM AI Aptitude Test as an objective measure by having workers who reported using AI in some capacity complete a short, structured assessment consisting of eight questions designed to evaluate both conceptual understanding and practical application. Responses on the SHRM AI Aptitude Test were scored using a frequency-based weighting approach to create standardized scores and minimize the impact of lucky guesses on final scores. These scores were then categorized into four groups according to their distance from the median score:

• Low: scores more than one standard deviation below the median.
• Medium-low: scores within one standard deviation below the median.
• Medium-high: scores within one standard deviation above the median.
• High: scores more than one standard deviation above the median. 

For workers with AI in their workplaces, an average of 46% of their work involves AI assistance. However, the average varies widely across job levels. Individual contributors reported that 34% of their work involves AI assistance, compared to 50% for managers and 63% for directors and above. These results indicate that directors and above lead AI adoption at their organizations and the nature of their responsibilities provides many opportunities for AI to support how that work gets done. 

Overall, workers reported that they use AI throughout typical work stages, such as planning, research and analysis, execution, review and quality assurance, and reporting. Workers were more likely to indicate that they use AI during the research and analysis stage, such as to augment gathering information, reviewing trends, and documenting incidents. This data suggests that workers are finding ways to incorporate AI into various stages of their workflows and they are primarily using AI to augment their capabilities rather than replace core responsibilities. 

As AI adoption increases, a new byproduct is AI-created content that is low-quality, derivative, and filled with factual discrepancies. Colloquially known as “AI slop,” this poor content is defined by its poor quality and, often, sheer quantity. AI can generate exponentially more material than human workers, but it also raises serious concerns about quality and originality in everyday work.

Higher AI use is linked to more reports of lower-quality, AI-assisted output. Overall, 41% of workers use AI at work, and about 44% of those users said their output includes “AI slop.” Because directors and above use AI more than others (63% versus 50% of managers and 34% of individual contributors), they were also more likely to report producing lower-quality AI output (31% versus 26% overall).

Most workers believe that using AI tools does not harm their ability to work independently. In fact, 68% stated they could resume their normal duties without delay if AI vanishes tomorrow. Even though 49% now rely on AI for tasks they used to do alone, 42% said they still feel confident completing specific jobs without AI assistance. However, when average scores are combined across the five AI reliance items, they indicate that workers rely more on AI than their own skills to accomplish tasks.

Although dependence on AI is increasing, workers strongly agree that human judgment matters now more than ever. In fact, 71% said critical thinking is more important now because of AI, and 63% feel these tools push them to analyze information deeply. And workers know the risks of AI use: 69% reported that you need strong critical thinking to use AI effectively, and the same number warned that lacking this skill can lead to bad decisions. 

Overall, when workers were asked to rank the top skills for an AI-driven work environment, they placed AI technical literacy first, followed closely by critical thinking. Meanwhile, communication, emotional intelligence, and adaptability were not as immediate of skills concerns for an AI-driven work environment. These findings suggest that while AI can handle routine work, critical thinking remains our most valuable asset. 

Those at the director level and above view AI differently than other workers. Compared to individual contributors and managers, they reportedly place a much higher value on critical thinking when using AI. However, they also said they rely on AI tools more heavily to complete their daily tasks. Interestingly, when these senior leaders were asked about ethical dilemmas, they leaned toward trusting AI technology over their instincts. If an AI recommendation conflicts with what they believe is the most ethical choice, 55% of directors and above said they would follow the AI's advice, while 45% would trust their own judgment (compared to 32% of individual contributors and 41% of managers who said they would trust their own judgment). As AI continues to evolve, it will be critical for leaders to strike a balance between leveraging technology and maintaining their own critical thinking and ethical standards.

Looking at differences between how workers rated their reliance on AI for tasks, those in the high aptitude group reported significantly lower reliance on AI than those in the medium-low group. Workers who have greater familiarity with and understanding of AI may use it with more caution, very aware of its potential biases and limitations, and therefore place greater emphasis on human oversight and evaluation. Conversely, lower-aptitude workers may be more inclined to view AI as authoritative and trustworthy, which could contribute to higher reliance on it.

Analyses of the perceived importance of critical thinking also reveal significant differences across aptitude groups. High-aptitude workers reported the highest levels of perceived importance of critical thinking, at a rate that is significantly higher than medium-low and low aptitude workers. This suggests that workers with higher AI aptitude are more likely to critically evaluate AI-generated outputs rather than rely on them at face value. It can be deduced, then, that higher-aptitude workers may be better equipped to recognize the limitations of AI systems and apply independent judgment when integrating AI into their work.

Nearly half of workers (46%) said they believe AI can perform a large portion of their current tasks, and 41% said the technology could handle their role with minimal human oversight. Furthermore, 42% expect their positions to change significantly as AI adoption grows. Despite this anticipated change, 7 out of 10 workers (69%) feel secure in their role. This feeling of job security is not unwarranted — SHRM’s Automation, Generative AI, and Job Displacement Risk in U.S. Employment data brief found that only around 6% of U.S. employment is at least half automated and has no definitive nontechnical barriers to automation displacement.

Clearly, workers recognize the power of AI to automate key aspects of their jobs. However, they do not fear widespread displacement. Instead, the opposite perception emerges, in which workers believe their overall roles remain secure even if AI automates specific tasks. This mindset highlights a clear distinction between job transformation and job replacement.

Managers provide a more specific perspective on how AI affects leadership. Most managers (64%) agreed that core leadership duties remain difficult to automate. This is likely due to viewing inherent human qualities that AI cannot replace — such as complex judgment, interpersonal skills, and decision-making — as essential management traits.

At the same time, managers acknowledge that AI can support their daily work. Over 3 in 5 managers (63%) stated that AI could automate certain decision-making tasks. This shows they welcome AI as a tool to enhance productivity rather than act as a full replacement. 

Despite this openness, structural concerns still exist. A slight majority (58%) agreed that AI might reduce the need for some management positions in the future. This creates tension within management ranks. Managers as well as directors and above view leadership as an established human function, but they also recognize that the increased efficiency from AI could reduce the total number of leaders needed.

Just under half of workers at large (45%) and extra-large (42%) organizations reported that their organizations have undergone a functional restructuring within the past 18 months. This compares to 39% of workers at midsize organizations and 31% at small organizations. When examining how these changes are communicated, AI frequently emerges as part of the narrative, both directly and indirectly. Among those who experienced restructuring, roughly half (45%) reported that it was directly attributed to AI adoption. Given the expected impact of AI, it may also be driving other communicated reasons such as efficiency improvements (54%) and cost reduction (44%).

Even when AI is identified as a contributing factor, the nature of its impact on work is more nuanced than job displacement. Among those who attributed restructuring to AI, the most reported outcome is the automation of specific tasks (58%), rather than the elimination of entire roles (31%). That being said, workers reported that nearly a third of AI-based restructuring efforts (31%) contributed to job loss.

Other forms of workforce adjustment are also prevalent when AI is communicated as the reason for organizational restructuring, including reduced hours (42%) and hiring freezes (32%), suggesting that organizations are shrinking. Organizations could be responding to broader pressures that often accompany contraction and market instability. This interpretation is reinforced by the finding that nearly half of CHROs (49%) identified economic uncertainty as a major challenge heading into this year.

Therefore, AI is less often a singular disruptive force and more often a catalyst embedded within broader organizational context and strategies around efficiency and productivity. Workers may perceive AI as the primary driver reshaping the labor market, but the surrounding context points to a more complicated reality: AI may, at times, function as a convenient explanation for restructuring decisions that are also being driven by economic uncertainty and organizational downsizing. There is more to explore beneath the surface of these organizational restructuring efforts.

Meanwhile, workers with postsecondary education in fields such as health and medicine (36%), social sciences (37%), and arts and humanities (34%) are less likely to anticipate the need for reskilling. It may be that the perceived need to reskill is highest in fields where AI is already more visible. In technical domains, where AI tools are rapidly advancing and being actively deployed, workers may feel more immediate pressure to adapt.

All academic disciplines were  viewed as more vulnerable to AI by workers who studied that field compared to those outside of it, with statistically significant differences in five of seven fields. On average, workers rated their own field as 14 percentage points more likely to become redundant due to AI than workers who did not study that field. This aligns with SHRM’s Automation, Generative AI, and Job Displacement Risk in U.S. Employment data brief, which detailed that automation potential varies greatly among different occupational groups, with computer and mathematical occupations identified as having the highest likelihood of more than half of their tasks becoming automated.

In each of these areas, those with direct experience or training in the field were far more likely to anticipate disruption, suggesting a more distributed or less concentrated expectation of risk overall. This dynamic points to an important psychological and experiential factor because proximity to a field appears to increase awareness of its vulnerabilities. Those leveraging the skills from their discipline may have greater visibility into which tasks are repetitive, automatable, or already being augmented by AI, leading to higher perceived risk compared to outsiders.

Most workers believe AI will ultimately increase the value of their field, but early-career professionals stand out for their heightened sense of risk and uncertainty. Having only recently entered the workforce, this group appears to be grappling most directly with the possibility that the skills they’ve just invested in may rapidly decline in demand.

Among entry-level to early-career workers, 27% believe AI will decrease the value of their field, more than double the share of advanced professionals at 13%. At the same time, they are less likely to express strong optimism, with fewer expecting AI to “significantly increase” the value of their field compared to their more-experienced peers. With limited experience but recent academic training, they are often closest to the foundational tasks and skills that AI is beginning to augment or automate. This concern cascades on top of an already-weakening labor market for college graduates, as SHRM’s Labor Force Snapshot: Recent College Graduates in the U.S. Labor Force detailed, because for the seventh year in a row, the 12-month average unemployment rate of recent college graduates surpassed the overall unemployment rate.  

About half of workers (51%) identified their organizational culture type as “Growth Collaborator,” in which individuals and teams are empowered to adapt, innovate, and take ownership. In these cultures, strong collaboration and open structures are esteemed. The next most common culture, “Disciplined Achiever,” accounted for 14% of workers. This culture type emphasizes consistency, productivity, and reliability through formal, centralized structures. The other six culture types were much less common, representing 5% to 9% of workers .

Given that Growth Collaborator was the most common culture type among workers’ organizations, most workers (57%) also indicated they have an open culture toward AI integration in their workplaces. Workers were more likely to agree that their organizations embrace AI technology as part of their future strategy, with a “learn as we go” mindset for integrating AI into processes and co-workers openly sharing what they learn about how to use AI tools. The remaining 42% of workers indicated a more closed culture around AI, with more hesitance to embrace how AI helps their business, a structured learning process for adopting AI into processes, and co-workers less willing to discuss how they are using AI. 

With an open and accepting culture for AI integration in organizations, workers are also more likely to take an enthusiastic, opportunistic attitude on AI tools. In particular, most workers were likely to report that they enjoy experimenting with using AI in their work, they are willing to experiment using AI in their work, and using AI allows them to focus on the parts of the job they prefer. Average scores of 4.0 out of 5.0 indicate that directors and above are significantly more likely to take initiative and experiment with AI in their work than managers (average of 3.8) and individual contributors (average of 3.4). 

These findings show that organizations with open, growth-focused cultures lead the way in AI adoption and experimentation. Teams that feel supported trying new tools and learning collaboratively are more likely to see positive results and embrace the technology’s potential. Prioritizing building a culture that balances innovation with sound guidelines ensures teams feel confident using AI to drive business value.

Our findings also show that workplace environment and culture play major roles in shaping these outcomes. Workers reported higher engagement when their organizations take an open approach to AI integration, build trust in leadership and the organization, and place a high value on critical thinking, but also rely less on AI for skills replacement. Two out of three workers (67%) who trust their organizations and leadership to act with integrity and consistency regarding the organizations’ mission and values were more likely to have higher engagement with their jobs compared to workers (6%) who do not have trust in their organizations.

Workers with strong organizational commitment reported greater value for critical thinking when using AI, higher overall trust in their organization, greater openness to embracing AI in their workplace culture, and less reliance on AI in their work. Interestingly, a more experimental and proactive attitude toward AI use is also linked to stronger organizational commitment, but it does not appear to relate to job engagement in the same way. This suggests that workers may value working for organizations that support innovation, even when that mindset does not directly change how engaged they feel in their daily work.

These findings show that AI culture matters as much as AI adoption. Organizations that foster trust, encourage thoughtful AI use, and reinforce the importance of human judgment are more likely to sustain an engaged and committed workforce. AI strategy should be viewed as part of workplace culture, not just technology deployment. When workers trust their organizations, feel supported in using AI responsibly, and believe their own skills still matter, then engagement and commitment are more likely to remain strong.