Artificial intelligence was sold to office workers as a way to clear inboxes, shorten meetings and finally claw back some time. Instead, a growing body of research suggests the first wave of AI adoption is piling on responsibilities, stretching workdays and, in many cases, eroding pay and bargaining power. The promise of smarter tools is colliding with the reality of lean budgets and aggressive automation targets, leaving many people doing more, not less, for the same money or even less.
I see a widening gap between the upbeat rhetoric around productivity and the lived experience of employees who are now expected to supervise algorithms, fix their mistakes and absorb the stress when things go wrong. The question is no longer whether AI will change work, but whether the way it is being rolled out will deepen burnout and inequality unless employers, HR leaders and policymakers change course.
The automation paradox: AI that creates more work
The first warning sign is what some researchers describe as an automation paradox, where tools that are supposed to simplify tasks end up making human jobs more complex. Instead of eliminating drudgery, early deployments of generative systems are shifting workers toward edge cases, judgment calls and exception handling, which are cognitively heavier and harder to schedule. One recent report warns that AI may “paradoxically increase the knowledge worker’s burden of handling complex tasks while simultaneously creating new workloads that negate automation benefits,” a pattern that is already visible in customer support, compliance and software engineering roles that now revolve around reviewing and correcting machine output.
That same analysis links the paradox directly to health and safety, noting that the new workloads can intensify pressure and extend working hours in ways that are difficult to monitor or regulate. When AI handles routine steps, managers often assume the remaining work is lighter, even as it becomes more mentally taxing and emotionally draining. The report, cited by Jan, frames this as a structural risk rather than a temporary glitch, warning that without explicit safeguards, the technology will quietly ratchet up expectations on the humans left in the loop.
More complex responsibilities, flatter pay
Alongside the automation paradox sits a more blunt economic concern: job content is getting harder while compensation barely moves. As AI tools spread through offices, humans are now being tasked with AI management roles, from prompt engineering to quality assurance and risk triage. These responsibilities demand higher levels of judgment and technical literacy, yet they are often bolted onto existing job descriptions rather than recognized as distinct, better paid positions. In practice, that means a marketing coordinator or claims analyst can suddenly find themselves responsible for supervising powerful systems without any corresponding shift in title, authority or salary.
The pay dynamic is particularly stark when employers frame AI as making roles “easier,” even as the underlying accountability grows. Because some leaders perceive that automation has simplified tasks, they argue that salaries do not need to rise and, in some cases, can even be compressed relative to historical norms. Reporting on a new adoption study notes that “Humans are now being tasked with AI management roles, adding to the pressure. Because roles are being perceived as ‘easier,’ therefore salaries aren’t rising,” a pattern that undercuts the idea that productivity gains will automatically flow to workers. That tension is captured in a recent analysis of how AI is reshaping job design, highlighted by Humans and the telling phrase “Because roles are being perceived as ‘easier.’”
The Great Squeeze inside HR and corporate budgets
These pressures are landing in workplaces already squeezed by muted growth and tight budgets. HR leaders describe a “Great Squeeze” in which organizations are trying to cut costs, adopt AI and maintain engagement all at once, often with fewer people. With muted economic and job growth, many companies are freezing headcount while still rolling out new tools that require training, governance and change management. That work typically falls on HR, learning and development and frontline managers, who are themselves under pressure to justify their budgets and show quick returns on technology investments.
The result is a feedback loop where employees are asked to absorb more change with less support, which erodes trust and accelerates burnout. One detailed review of this trend, titled The Great Squeeze, warns that “With muted economic and job growth, organizations are increasingly turning to AI and automation while simultaneously reducing budgets, leading to higher workloads and lower levels of employee engagement.” That combination, “How HR Can Address Employee Burnout Amid Budget Reductions and AI Adoption,” underscores that the technology is arriving in a context of scarcity, not abundance, which shapes how its benefits and burdens are distributed.
Labor market impact: more churn, not instant job loss
Despite the anxiety, the near term labor market story is more nuanced than a simple wave of layoffs. Analysts tracking AI’s labor market impact argue that the immediate effect is more likely to be job reshaping and internal churn than mass unemployment. One recent overview of Labor Market Impact notes that “In the immediate term, expect more task-level automation and role redesign, often without formal company deployment,” as employees quietly adopt tools like ChatGPT, GitHub Copilot or Midjourney to speed up parts of their work. That informal experimentation can boost individual productivity, but it also makes it harder for organizations to track where skills are shifting and which roles are becoming structurally dependent on AI.
Over time, the same report, framed as “Understanding Workforce Transformation,” suggests that the bigger story will be how tasks are reallocated across roles rather than a clean swap of humans for machines. How that plays out will depend heavily on whether employers invest in reskilling and internal mobility or simply use automation to justify hiring freezes and outsourcing. The authors emphasize that “How is AI reshaping work?” is not a rhetorical question but a practical one, with different answers in software development, logistics, healthcare and marketing. Without deliberate planning, the default path is that workers absorb the complexity of hybrid workflows while the financial upside accrues elsewhere.
ROI problems and skipped change management
One reason workers are feeling squeezed is that many organizations are chasing AI headlines faster than they are building the foundations needed to use the tools responsibly. New research on enterprise deployments finds that success with workplace AI will not come from adding more tools, but from rethinking processes, governance and training. Looking ahead to 2026, analysts argue that the companies that actually see returns will be those that slow down enough to redesign workflows, clarify accountability and invest in human skills, rather than simply layering chatbots on top of existing systems.
Yet a significant share of early adopters have largely skipped those steps, rolling out pilots without robust measurement or employee input. One synthesis of these findings notes that “Looking ahead to 2026, these findings paint a picture” in which the most successful organizations are those that treat AI as a strategic capability, while “a large share of AI adopters largely skipped” the basics of change management and stakeholder engagement. That gap is highlighted in a recent industry roundup that points to Looking at AI’s ROI problem as a governance issue as much as a technical one. When organizations rush, the burden of figuring out how to use new systems “most effectively” falls on individual workers, who must improvise their own guardrails and workarounds.
The AI wage premium, and who actually gets it
At the same time, there is real money on the table for a narrow slice of the workforce. The AI wage premium is real, and it is reshaping salary bands in tech, finance and consulting. One widely cited analysis, titled The AI Wage Premium Is Real, puts the average salary bump for AI skills at 28 percent, drawing on data from Exploding Topics, 2025, and Forbes, 2025c. Let that figure sink in: workers who can credibly demonstrate expertise in machine learning, prompt design or AI product management are commanding nearly a third more pay than peers in similar roles without those skills.
However, that premium is not evenly distributed, and it does not negate the broader trend of more work for flat pay. The same analysis, framed as “AI Labor Trends 2026: What’s Really Happening with Jobs (And Why Let),” suggests that the biggest gains are concentrated among senior engineers, data scientists and product leaders in large firms, while frontline staff in customer service, operations and content production see their tasks automated without a corresponding raise. Exploding Topics and related datasets show that the 28 percent figure is an average across roles and industries, which means some workers are seeing far larger jumps while others see none at all. For many employees, the arrival of AI looks less like a personal windfall and more like a new performance bar they are expected to clear just to keep their current job.
Marketing as a test case for “operationalized” AI
Marketing is emerging as one of the clearest test beds for how AI can both empower teams and intensify expectations. In that field, leaders are already arguing that 2026 belongs to the teams that operationalize AI in a disciplined way, rather than treating it as a side project. The most successful marketing organizations are expected to embed generative tools into campaign planning, creative development and analytics, using them to personalize content and scale campaigns across channels. That shift is not theoretical; it is showing up in job postings that ask for experience with tools like Jasper, HubSpot’s AI assistants and Meta’s Advantage+ suite alongside traditional skills in segmentation and copywriting.
Yet the same playbook that promises efficiency can also raise the bar for output and responsiveness. When AI can generate dozens of ad variants in seconds, managers may expect more testing, faster iteration and round the clock optimization, often without adding headcount. A recent trend piece, featuring Jasper’s Loreal Lynch, argues that “2026 Belongs to the Teams That Operationalize AI In 2026, the most successful marketing organizations won’t treat AI as a shiny object, but as a core part of how they plan and scale campaigns across channels.” That vision, captured in a Belongs style manifesto, is compelling, but it also risks normalizing a pace of work that is only sustainable if organizations invest in training, guardrails and realistic performance metrics.
Will AI make work worse, or just different?
Underneath the statistics sits a more basic question: is AI going to make work worse, or simply different? Public sector technologists and policy experts are already wrestling with that dilemma as governments experiment with chatbots for citizen services, predictive tools for maintenance and AI-assisted drafting for legislation. Some argue that, used well, these systems can reduce drudgery and free up staff for higher value tasks like casework and strategic planning. Others worry that without clear boundaries, they will instead accelerate a culture of constant availability and data-driven micromanagement.
One recent discussion framed the issue bluntly under the heading Is AI going to make work worse, noting that even basic questions like which browsers to support, such as Chrome, Firefox or Safari, can shape who benefits from new tools. The same piece points out that legacy environments like IE 11 Not Supported can limit access to modern systems, creating a divide between workers who can tap AI assistance and those stuck with older software. As agencies and companies alike figure out how to train staff on new AI systems most effectively, the risk is that early missteps will harden into norms that treat constant monitoring and algorithmic nudges as just part of the job.
Wages that rise, then fall, as AI spreads
Economists are also warning that even where AI lifts wages, the effect may be temporary. A new Brookings Institution paper, summarized in recent coverage, argues that AI could initially boost pay for workers whose tasks are complemented by intelligent systems, only to drive those wages down as automation expands and competition intensifies. In the early phase, companies may pay a premium for employees who can integrate AI into workflows, design prompts or oversee deployments. Over time, as those skills become more common and more tasks are fully automated, the bargaining power of those same workers could erode.
The authors put it starkly: AI may lift wages, then crush them, particularly in occupations where “human workers become the bottleneck” for tasks that remain manual or judgment based. One summary notes that “A new Brookings Institution paper finds that AI could initially boost wages, but then drive them down as automation expands,” highlighting the risk that the technology will make the economy more productive while leaving many individuals worse off. That dynamic is captured in coverage on Brookings, which emphasizes that the long term impact depends on how gains are shared and whether policy keeps pace with technological change.
When humans become the bottleneck
The same Brookings analysis, echoed in financial reporting, drills into a subtle but crucial shift: as AI handles more routine work, the remaining human tasks can become the limiting factor in production. In sectors like legal services, accounting and software development, that might mean that drafting, coding or initial analysis is largely automated, while review, client interaction and final sign off remain human. If organizations treat those remaining steps as interchangeable or easily outsourced, they may feel less pressure to raise wages, even as the responsibility and risk attached to those tasks grows.
Financial coverage of the Brookings findings notes that “Brookings says that AI could initially boost wages, but then drive them down as automation expands. AI may make the economy more productive, but it could also reduce the bargaining power of workers whose tasks are partially automated, especially when human workers become the bottleneck.” That phrase, “human workers become the bottleneck,” is doing a lot of work. It suggests a future in which the scarcest resource is not data or compute, but human attention and judgment, yet the market may still undervalue those qualities. The tension is captured in a detailed summary on Brookings that urges policymakers to anticipate these shifts rather than reacting after wage compression has already taken hold.
What needs to change to avoid “more for less”
All of this points to a simple but uncomfortable conclusion: without deliberate intervention, AI adoption is likely to deepen existing workplace strains, not relieve them. To avoid a future where workers consistently do more for less, employers will need to treat AI as a redesign challenge, not just a cost cutting tool. That means mapping which tasks are being automated, clarifying who is accountable for AI outputs, and adjusting job descriptions, staffing levels and pay scales accordingly. It also means involving employees in decisions about where and how to deploy tools, rather than imposing systems from above and expecting frontline staff to absorb the fallout.
From my perspective, the most promising path is one where organizations pair AI investments with equally serious commitments to training, mental health and fair compensation. HR leaders grappling with The Great Squeeze have a chance to set new norms that recognize AI management as skilled work, worthy of recognition and reward. Policymakers, informed by analyses from Brookings and others, can update labor standards and social protections to reflect a world where “human workers become the bottleneck” in ways that are both economically and ethically significant. If they do not, the headline warning that AI adoption may leave workers doing more for less pay will not be a cautionary tale, but a baseline description of the new normal.
More from Morning Overview