EY refers to the global organization, and may refer to one or more, of the member firms of Ernst & Young Global Limited, each of which is a separate legal entity. Ernst & Young Global Limited, a UK company limited by guarantee, does not provide services to clients.
Recent Searches
Can AI advance toward value if workforce tensions linger?
Reaping value from AI investments requires mastering the tensions between talent and tech or risking an AI productivity drop of 40%.
In brief
- AI usage at work is widespread (88%), yet only 28% of organizations have positioned employees to achieve transformative business impact from AI.
- AI Advantage needs toolset, skill set and mindset investment while managing key tensions, including bringing your own AI behaviors and talent attrition.
- Leaders build Talent Advantage by excelling in talent health, learning culture and strategic rewards that amplify AI adoption excellence.
CEOs across industries now see artificial intelligence (AI) as essential, not optional. Yet, according to a recent MIT study, despite billions invested in generative AI (GenAI), roughly 95% of organizations report no measurable return, while only about 5% of integrated AI pilots are extracting millions in value and demonstrating meaningful P&L impact.1 This, despite companies’ massive investments in AI – a third (35%) of senior leaders whose organization is investing in AI anticipate spending US$10 million or more next year, according to a recent EY study in the US.2 As Gartner© observed in March 2025, “Despite the excitement surrounding AI, its impact on productivity has been inconsistent, leading to what some describe as the AI productivity paradox.”3
The EY Work Reimagined study has tracked the evolution of work for years. This year, building on our past work on Talent Advantage, we zeroed in on AI in the workplace – specifically, why some organizations achieve transformational results while most see only modest gains. We surveyed 15,000 employees and 1,500 employers across 29 countries and looked at a range of outcome measures including productivity, quality of work, decision making and various dimensions of the work experience. We looked at how employees are adopting AI and the benefits at work and then compared that with the benefits companies are seeing from a business perspective.
The findings of our Work Reimagined 2025 study reveal the scale of the challenge. Nearly nine out of 10 employees now use AI at work, yet only 28% of organizations are positioned to turn AI deployment into high-value outcomes. Our research shows why: employees may be saving a few hours here and there but nothing that fundamentally changes how work gets done or how the business performs.
What separates the 28% from everyone else? The organizations achieving transformational results have honed five strategic capabilities working in concert: having the right approach to recruiting and retaining talent; driving AI adoption at scale; building continuous learning into daily operations; reshaping culture and workplace norms; and aligning rewards with new behaviors and outcomes.
AI investment on its own isn’t enough. When new technology lands on fragile talent foundations – weak culture, insufficient learning, misaligned rewards – productivity benefits lag by over 40%.
Creating the Talent Advantage requires five key dimensions
1
Chapter 1
How to drive AI Advantage – and why it’s not enough
Widespread adoption masks shallow use and fragile talent foundations.
Employee usage of AI at work is now widespread, with 88% of survey respondents saying they use AI at work to some degree and 37% use it daily. Usage spans from 80% of essential workers to 94% of knowledge workers, with near-universal adoption among leaders and managers.
Mind the maturity gap
However, this surface-level adoption masks a deeper problem. While most workers use AI for basic tasks like searching for information (54%) or summarizing documents (38%), only 5% qualify as advanced users who blend multiple tools to unlock roughly a day and a half of additional productivity per week. Advanced users extract far more value, using AI as a thought partner rather than a simple tool.
Skill sets, toolsets and mindsets drive the AI Advantage
What drives this gap between basic AI adoption and true AI adoption excellence is a model we call the AI Advantage. We analyzed more than 15 variables in our 15,000-employee dataset to understand what fuels employees who see the most AI adoption value at work, a measure of adoption success from 0-100 that examines median hours unlocked per employee per week as an outcome variable. We found that three interconnected drivers – skill set, toolset and mindset – power this AI Advantage.
Skill sets unlock productivity
Unsurprisingly, AI training is key to success, accounting for nearly 50% of the AI Advantage score. Employees with 81 or more hours of AI training per year save 14 hours per week, compared to just 3 hours for those with fewer than 4 hours of training. Yet only 12% of surveyed employees received this level of training in the past 12 months, which explains why the global AI adoption value score sits at just 34 out of 100. Manager proficiency also plays a crucial role – employees who are confident their managers know how to use AI effectively see better outcomes. However, the catch is that these highly skilled employees are also 55% more likely to leave. Employers need to align the employee value proposition for employees with hot skills in the market. It is critical to grant continued access to the latest technology as well as opportunities to translate the skills to career options.
Toolsets provide the right tools for the right role and clear guardrails
The right AI tools matter, and employees who report their AI tools are tailored to their specific role also demonstrate significantly higher adoption value. However, in parallel, employees who are most passionate about AI potential are also taking matters into their own hands with 23% to 58% of employees bringing their own AI to work, and even paying for their own subscriptions. Further, there is substantial pressure to perform, with 64% of employees perceiving a workload increase in the past 12 months and 38% fearing job loss without replacement. A new Harvard study agrees that some of these fears may be valid, with roles for junior workers, in particular, shrinking sharply since 2023.4 This makes it even more critical to provide employees with the right toolset and to create the proper guardrails for personal tool usage.
Mindset becomes the new multiplier
Mindset drives adoption. Having formal goals or incentives to adopt AI, along with involvement in organizational AI programs, correlates strongly with time saved and positive outcomes. Leading organizations are increasingly formalizing business and individual goals to adopt and scale AI and agents. They’re coupling upskilling programs with investments in culture, leadership and engagement to shift behaviors and drive impact. When employees see AI adoption as part of their performance expectations and organizational culture, they engage more deeply. Employers with Talent Advantage have 63% reporting that culture is significantly better than 12 months ago vs. 22% in the Talent Middle and 3% in the Disadvantage group.
Organizations that build an AI Advantage face new tensions
However, getting these drivers right is necessary but not sufficient. Organizations that build an AI Advantage face new tensions that can offset productivity gains. Bridging the gap between AI investment and tangible impact requires a balanced approach that considers both technological capability and human readiness. The Talent Advantage – a framework of five strategic capabilities – can help to manage these tensions.
2
Chapter 2
Five capabilities that unlock high-value AI outcomes
Less than a third of organizations master the Talent Advantage that makes transformation enduring.
Organizations with weak talent foundations see productivity benefits lag by over 40% relative to organizations with what we’ve termed Talent Advantage – five interconnected strategic capabilities that create the foundation for sustainable transformation. Only 28% of organizations in our survey demonstrate strength across all five – not as separate initiatives but as an integrated system where each capability reinforces the others.
1. Talent health and flow
Talent health and flow drive everything else, measuring overall conditions for success using employees’ likelihood to recommend their employer to friends and family. Based on our analysis of the variables that predict net promotion (measured from 0-100 based on employees’ self-reported importance on the largest effect on the outcome to recommend the company), the global Talent Health score stands at 65 out of 100, with culture accounting for 44%, rewards 32% and development 24%.
Talent Advantage employers achieve dramatically higher talent health scores. While only 20% of employees at Talent Disadvantage organizations are likely to promote their company as a great place to work, 89% are likely to do so at Talent Advantage organizations. This creates network effects – employee promoters become talent magnets who attract more high-performers.
Talent flow remains foundational and even employees in sectors and at companies they promote will switch jobs. In 2025, the overall intent to quit has dropped to a four-year low of 29%, down from the “Great Resignation” peak of 43%. However, this masks an important dynamic: job jumping remains high for those with the most AI learning, with 45% quit intent among this group who continuously weigh internal and external career and pay opportunities. The paradox, which broadly surfaces with highly trained, high-value talent – building capability while increasing flight risk – becomes one of the critical tensions that organizations must navigate. Employers can address this risk by ensuring these critical employees have continued access to tech, career opportunities, and are well rewarded.
2. AI adoption excellence
AI adoption excellence requires frequent, sophisticated use of AI with role-specific tools and strategic goals. Organizations maximizing employee AI adoption value create opportunity to unlock eight to 14 hours per week of time savings as employees leverage higher complexity use cases. Their advanced users don’t just use AI more, they use it differently, treating it as a colleague, coach and thought partner rather than a simple automation tool. Top performers who adopt this collaborative mindset achieve gains more than twice as great as their peers.
As we realized our own custom AI tool, EYQ, would eventually not be able to keep up with the rapid pace of technology advancement, we pivoted to providing an alternative enterprise version with GPT-5 as well. This flexibility allows us to put employee needs first, while maintaining security and compliance.
However, organizations need to provide the right AI tools to get their employees to use the AI. Even with tailored tools, given pressures to perform, employees are often taking matters into their own hands. Between 23% and 58% of employees bring their own AI to work, with variation by sector. This “shadow AI” represents untapped innovation potential but also governance and security challenges that organizations must address systematically. “I think you have to be really clear around the sort of compliance issues the organization faces as a result of using some of these AI tools,” acknowledges Joe Depa, EY Global Chief Innovation Officer. “As we realized our own custom AI tool, EYQ, would eventually not be able to keep up with the rapid pace of technology advancement, we pivoted to providing an alternative enterprise version with GPT-5 as well. This flexibility allows us to put employee needs first, while maintaining security and compliance.”
3. Learning and capability development
Learning and capability development stands as the strongest predictor of AI success. Employees that receive the 81-plus hour threshold see transformational results. However, higher levels of training create the learning paradox: employees with more than 80 hours are 55% more likely to quit compared to the average. Advanced training makes skills more marketable and external markets may often reward AI expertise faster than internal promotion cycles can respond.
Talent Advantage organizations address this by pairing intensive learning with retention and rewards strategies and career development. They create internal talent marketplaces, design progressive skill certifications tied to tenure and build learning cohorts that develop peer networks because social capital increases retention. The proportion of employees receiving at least 80 hours of AI learning per year jumps from 15% at Talent Disadvantage organizations to 42% at Talent Advantage organizations.
4. Culture and workplace transformation
This kind of transformation provides the enabling environment for AI integration. Leadership vision and cultural alignment are essential for successful AI adoption. Culture scores have improved significantly, with 60% of employees now agreeing culture is significantly better than 12 months ago, up from 48% in 2021. Employers have made progress in closing gaps in ways of working, enabling team connections, and helping employees feel more trust and support.
Talent Advantage employers dramatically outperform on culture metrics, with 63% reporting culture is significantly better than 12 months ago versus only 3% of employees at Talent Disadvantage organizations. Culture accounts for 44% of the Talent Health score, driven by caring leaders, supportive employers, empowering managers, and team connection. AI adoption itself can strengthen these cultural elements when organizations frame it as a collaborative learning experience rather than an individual technical skill.
5. Strategic total rewards
Strategic total rewards must be personalized and flexible, aligning with evolving employee needs and AI-driven roles. Unlike in the past, where employees focused primarily on career opportunities inside their organizations, employees with AI skills can easily look outside for their next role. These skilled workers become more focused on a work experience with great technology, flexibility and opportunities for growth that lead the market.
Talent Advantage organizations understand that rewards drive approximately 32% of talent health and they excel at making total rewards flexible. Fully half of Talent Advantage employers strongly agree their rewards meet employees’ needs, compared to just 7% of Talent Disadvantage organizations.
3
Chapter 3
Navigating the five critical tensions to reach the Talent Advantage
Leading organizations don’t avoid tensions; they actively manage them.
Every organization that pursues a Talent Advantage must pay attention to five important tensions that can derail progress. The key is to lean into these challenges, providing the right support at the right time while learning as you go. Here are five tensions along with the key actions leaders can take.
1. Address the learning-retention dilemma
The 81-plus hours of AI training that drives adoption excellence also makes employees 55% more likely to leave. Employees with fewer than 4 hours of AI learning have 21% quit intent, rising to 45% for those with 81 or more hours. The motivations of highly trained employees shift too. Those with more than 40 hours prioritize opportunities to work with the latest technology and enhanced flexibility over traditional compensation and career advancement.
Leaders face a dilemma: under-invest in learning and never build AI capability or invest heavily knowing it increases flight risk. The answer lies in learning, rewarding and connecting to career paths as you go – pairing intensive training with evolved retention approaches that match how AI-skilled employees now think about their careers. Build internal pathways that showcase opportunities for growth and technology access before employees look externally. Further, create learning experiences that also build social capital and peer networks, recognizing that relationships anchor people as much as roles do. Critically, calibrate the total rewards strategy to reflect what AI-skilled workers actually value, including access to cutting-edge technology, flexibility in how and where they work, and opportunities for continuous skill development, not just traditional pay increases.
2. Maximize AI-enabled time gains
AI delivers tangible time savings, with employees reporting an average of eight hours saved per week. However, many organizations struggle to translate these efficiency gains into meaningful business transformation. Leading companies in sectors like Wealth Management, Technology, and Banking are saving even more time (10 to 12 hours per week), but the critical question remains: how should this newly available capacity be strategically deployed? JPMorgan Chase offers a compelling example. The bank deployed LLM Suite – a generative AI tool – to 50,000 employees across its Asset & Wealth Management division. The tool performs work equivalent to a research analyst, but rather than simply banking efficiency gains, JPMorgan is strategically redeploying analyst capacity toward higher-value work: complex interpretation, strategic advisory and deeper client relationships.5
The key to avoiding the productivity trap is being deliberate about where saved time gets reinvested and recalibrating how ROI is measured. As EY Global Chief Innovation Officer Joe Depa notes, “the value of AI has three key components: productivity, quality and efficiency. Organizations fixated solely on hours saved miss critical dimensions like improved accuracy, reduced error rates, and enhanced decision-making quality.”
Redesign roles alongside AI deployment. Determine what employees should stop doing, what high-value activities they should focus on, and how they can elevate their contributions. Establish clear expectations about allocating time savings between strategic growth initiatives and creating space for innovation, learning and adaptation. Consider piloting this approach in one function and measure both productivity and employee engagement metrics. Make role redesign a standard component of AI implementation planning rather than an afterthought. Additionally, create feedback mechanisms to continually refine how AI-enabled time savings translate into higher-value work, to more effectively gauge ROI and allow efficiency gains that drive meaningful transformation.
The value of AI has three key components: productivity, quality and efficiency. Organizations fixated solely on hours saved miss critical dimensions like improved accuracy, reduced error rates, and enhanced decision-making quality.
3. Deal directly with the anxiety-innovation gap
While organizations struggle to find talent, 38% of employees fear job loss due to AI. The same proportion, 38%, worry about overreliance on AI eroding human skills, expertise and learning. This fear coexists with innovation demands. Organizations need employees to experiment with AI and reimagine their work. However, workforce anxiety creates hesitation, resistance and defensive behavior that protects current roles rather than transforming them.
Embrace emotions rather than ignore them. Leaders must articulate a clear AI vision that addresses workforce concerns with a top-down commitment to managing the stresses of transformation – the increased workloads, the fears about job security, the anxiety about obsolescence.
Communicate not just what AI will do, but what humans will do that’s more valuable. Identify employees whose roles have been elevated by AI and document their stories. Specifically, what changed and why their work matters more now. Share these through the organization’s communication channels.
Give employees voice in how AI gets implemented in their areas, allowing those closest to the work to help shape its future. Additionally, consider building programs that develop new skills before old ones become obsolete, making continuous adaptation the norm.
4. Resolve the shadow AI challenge
Between 23% and 58% of employees bring personal AI tools to work. Stifling this kills innovation, but ignoring it creates security, governance and compliance nightmares. Enterprise tools lag behind consumer AI experiences, and employees won’t wait for approval when they can solve problems immediately.
Survey your workforce to understand what personal AI tools they’re using and why and be sure to incentivize disclosure rather than punishing it. Then, create “AI sandbox” programs that channel this energy into governed experimentation. As Tyler Buffie, EY Global Data and AI Strategy Leader put it, “employees need time and space to practice with the AI tools, figure out how they can use them and then apply it to their workflows.” Give employees explicit space to test personal tools under clear parameters. Fast-track promising solutions into enterprise adoption to create a reverse innovation pipeline. Turn shadow AI users into internal champions and tool scouts. Establish clear governance boundaries while defining “innovation zones” for controlled risk-taking.
5. Energize reorganization to combat fatigue
Roughly eight out of 10 Talent Advantage employers have already significantly reorganized due to AI, yet 74% recognize this still needs to evolve. Leaders must continue changing to capture AI value, yet constant restructuring exhausts the workforce. AI transforms work every quarter, but traditional reorganizations take 12 to 18 months.
To help address reorganization fatigue, create autonomy for parts of the organization to experiment with new structures in short cycles. Assign clear roles and responsibilities while delegating decision-making authority appropriately. Establish which parts of the organization will remain stable alongside which areas need to transform, giving employees certainty about what won’t change. Build change capacity as an ongoing capability rather than treating each reorganization as a singular traumatic event.
Map your organization into stability zones and transformation zones, then communicate this clearly. Enable the people executing these experiments to have real authority to make decisions about how work gets organized, not just permission to suggest changes that require endless approvals.
Shape the AI and workforce future with confidence
Only 28% of organizations have reached Talent Advantage, unlocking transformational results. The remaining 72% have a choice: become fast followers who quickly build the five capabilities while navigating the tensions or fall further behind as the gap between leaders and laggards widens.
To achieve the right mix of five strategic capabilities and reach the Talent Advantage, organizations must orchestrate competing priorities: learning and retention, innovation and security, global and local talent needs, centralized vision and distributed execution. This requires systematic excellence, not heroic individual efforts.
Talent Advantage matters because it predicts not only productivity but also whether organizations can achieve strategic goals beyond productivity, including enhanced performance, better decision-making, improved employee wellbeing and culture. In a fast-changing business environment, where an adaptable and resilient workforce is essential, organizations with Talent Advantage have the opportunity to pursue ambitious growth while managing inevitable talent tensions. Those without get stuck optimizing existing processes while competitors reimagine business models.
Sustainable advantage in the AI era depends on combining the ability to build strong human foundations and advanced technology. There’s no question whether AI will transform an organization’s industry and workforce. It will. The question is whether organizations will shape their transformation – or react to it.
The five capabilities provide what to build. The five tensions reveal the challenges organizations will face. The 28% who have achieved Talent Advantage prove it’s possible. Now the question becomes: will your organization build the systematic excellence required to join them?
You can write it, or you can read about it. The choice is yours.
Summary
The 2025 Work Reimagined survey reveals that while 88% of employees use AI, only 28% of organizations achieve transformational results. The difference? These leaders don’t choose between AI and people – they orchestrate five strategic capabilities to amplify each other while leaning into tensions everyone else avoids: learning that drives retention risk, productivity that increases workload, innovation that sparks anxiety. Sustainable advantage requires both advanced technology and strong human foundations, not one or the other.