AI brings more work, not less
„AI imposes hidden burdens on workers through error correction, oversight, and low-quality content, while company-level productivity lags.”
Counter-narrative
AI saves time and is increasingly appearing in corporate financial results. This is reflected by McKinsey (37% of companies feel the impact), US Census time-saving data, and AI company CEOs promising leisure time.
Carriers
NBER survey, BCG, Gartner, Glean, Deloitte, CHRO surveys (HR Dive), New York Times
Analysis
The narrative describes the gap between the productivity promise and everyday experience. AI usage is spreading, but saved time is minimal or absorbed by fixing errors and supervision. The organization does not become more efficient, and work assessment becomes harder. Gartner data fuels this: according to HR leaders, 95% of organizations have adopted AI, but only one-fifth have seen significant results. Glean data points in the same direction regarding burnout and turnover risks, as well as managerial tasks performed without pay raises. This benefits employees and change management consultants, as it urges redesigning workflows. Its weakness is that implementation is early, and the burden may partly be a temporary learning cost.
What's at stake: Whether AI investments translate into productivity or are consumed through worker overwork.
Narrative sub-threads
AI waste and bug-fixing burden
Fixing AI errors, monitoring outputs, and cleaning up low-quality AI content consume measurable work time. Surveys indicate this takes more time for many workers than actual productive work.
This branch targets the output itself: humans must verify and correct AI-generated errors and 'slop'. It is not work organization or leadership roles that are changing.
→ lasting · evidence: weakLonger working hours, invisible extra work
The freed-up time must be filled with more work; Meta's head said so explicitly. Employee reports indicate that AI use goes with longer working hours, a heavier workload and burnout.
The mechanism here is increasing workload and pace, not fixing faulty output. This is supported by worker reports and CHRO opinions.
→ lasting · evidence: moderateBurden shifted to managers and HR
AI implementation places new tasks on managers and HR. More subordinates must be supervised, evaluating AI-generated work is harder, and HR administrative burdens do not decrease. More and more people perform managerial duties without a pay raise.
The actors are the manager and the HR professional; the burden appears in evaluation, oversight, and administration, not with the employee doing the work.
→ lasting · evidence: strongLagging company-level productivity
Despite widespread AI adoption, most leaders see no measurable corporate productivity or financial results. According to Gartner, only one in five HR leaders saw significant results; the NBER reports that 90% of leaders see no productivity impact; and McKinsey finds only 37% of companies feel the effect in financial results. Research shows that layoffs citing AI further worsen the performance of those who remain.
This strand relies on organization-level return data (Gartner, NBER, McKinsey, Deloitte). It measures company-level results, not individual workload.
Sub-thread stories
AI waste and bug-fixing burden
Glean: AI creates hidden workload for Australian employees, with 6.5 hours a week spent on oversightemployeeTell your manager how much time you spend checking AI so workloads can be planned.HR professionalTrack the AI oversight burden alongside burnout and turnover indicators.Freelancers are hired to clean up 'soulless' AI-generated contentemployeeCleaning up AI content can become a new market niche for freelancers, with their own pricing.Shopify CEO says AI-proof-of-need rule led staff to throw "slop grenades" at each othermanagerQuality rules should accompany the expectation to use AI.employeeHanding over an AI draft without checking it burdens colleagues.Skill erosion and a 'verification tax' emerge as side effects of AI at workmanagerJuniors should deliberately be given the chance to learn tasks manually, and time for checking AI output should be built into plans.early-career professionalTo gain experience, it is worth practising the basics without AI as well.employeeChecking AI output is new work that should also be reflected in performance reviews.Companies that rushed into AI are now pulling back as errors and poor quality mountmanagerMeasure the error rate and rework in AI-replaced processes before making headcount decisions.employeeChecking and correcting AI output is valuable work, so make it visible in performance reviews.Workers spend more time correcting AI errors than doing actual work, surveys findemployeeTrack and report the time spent checking AI output, as it determines the real productivity gain.managerInclude correction time in productivity metrics, otherwise the return on AI investment is distorted.Regular AI users work more: mental load and time spent fixing "AI slop" riseemployeeAI does not automatically reduce working hours; checking outputs also takes time.managerWhen measuring productivity, also measure time spent on corrections, and set limits against pressure to speed up.HR professionalMonitoring burnout risk and workload is a new HR task.
Longer working hours, invisible extra work
Tech leaders say AI means less work, while staff report working up to 90 hours a weekemployeeDespite AI promises, expect extreme workloads at tech companies, so clarify working hours during negotiations.HR professionalBurnout risk and turnover must be managed when the AI pace involves extremely long working hours.Zuckerberg says AI will shrink companies but not kill jobsemployeeAccording to the executive's message, time saved may mean more work rather than rest, so workload expectations should be clarified in advance.managerIf time saved with AI is used to raise performance expectations, account for the risk to satisfaction and of burnout.AI-based workforce management worsens working conditions for LanguageLine interpretersemployeeAI-driven scheduling can worsen working conditions.HR professionalThe human impact of algorithmic work organisation needs to be measured.Is AI making employees' workloads heavier?employeeInstead of time saved by AI, more work may end up on their desk.managerWorkload should keep being measured after AI is introduced.CHROs say AI creates 'invisible' workloadHR professionalWorkload assessments should also measure the checking and correction tasks that AI generates.managerProductivity expectations should be set with the extra work surrounding AI in mind.employeeIt is worth documenting the additional tasks linked to AI.AI in customer service: efficiency or burnout?employeeAutomation in customer service can also bring a heavier workload.managerAt a contact centre, the impact of AI should also be tracked through burnout indicators.
Burden shifted to managers and HR
AI puts new pressure on managers; half of CHROs doubt their managers are readymanagerSupport your team's AI use, as it also shapes how employees are perceived.HR professionalLaunch a dedicated programme to prepare managers for AI.employeeIf managers offer no support, ask for clear guidelines on AI use.AI makes staff more efficient, but companies lack clear goals and judging expertise has become hardermanagerAlongside AI use, set clear goals and metrics.HR professionalIn performance reviews, assess understanding and judgment as well as output.HR staff still spend half their time on administration despite AIHR professionalWorkflows must be put in order before adopting AI, otherwise administration will not fall.managerTo get a return on HR automation, invest in process design as well.AI is pushing more people into managerial work without a pay riseemployeeIf AI gives you managerial duties, raise this in role and pay negotiations.managerReassigning tasks without adjusting pay can breed dissatisfaction.HR professionalJob descriptions and pay must be aligned with the tasks that AI has changed.More than half of managers find it harder to evaluate staff work produced with AImanagerRedesign how goals and output quality are defined for AI-accelerated work, rather than measuring volume.employeeDocument which parts of the work involved human judgment, so your contribution is visible at review time.HR professionalRework the evaluation system for roles that use AI.
Lagging company-level productivity
Gartner: 95% of HR leaders use AI, but only one in five has seen significant resultsHR professionalAI training should also be planned for newly emerging roles, and its results measured.managerThe ROI of AI adoption needs to be measured, because usage does not equal impact.AI-linked layoffs hurt the productivity of the workers who remainmanagerAfter layoffs, managers should pay particular attention to remaining staff's trust and workload, or performance will drop.HR professionalCommunication and support after layoffs are tools for preserving productivity.90% of executives report no AI productivity gain, yet some firms keep cutting jobsemployeeWorkers should document overload after layoffs and report it to their manager.managerManagers should measure the real productivity effect of AI before using it to justify headcount cuts.HR professionalHR professionals should monitor the engagement and workload of staff who remain after layoffs.AI-driven layoffs may undermine the productivity gains companies expectmanagerBefore AI-based headcount cuts, the economics, the evidence and the reversibility should be tested separately.employeeRehiring often comes with lower pay, so it is worth thinking through your negotiating position in advance.HR professionalThe costs and loss of knowledge from rehiring after layoffs should be included when evaluating the plan.AI payoff only partly visible: McKinsey, PwC and TechCrunchmanagerReturns must be measured; costly deployment alone does not deliver results.HR professionalEmployees are using AI, but profit comes slowly; this is a planning risk.AI has not sped up recruitment, Hungarian reports sayHR professionalThe effect of AI tools on recruitment should be measured, because speed-ups are not automatic.managerThe return on recruitment AI investment should be tracked over time and with quality indicators.Companies spending more on AI are spending less on workers; productivity gains are individual, not company-widemanagerThe return on AI investment should be measured at the workflow level, not just in usage.employeeIndividual productivity gains may not translate into higher pay.Most UK workers use AI but save little time, Deloitte survey findsmanagerAI use alone does not deliver time savings; workflows must be redesigned too.employeeTime saved with AI is hard to measure, so apply it to specific tasks.
Google search interest
Hungary
| Topic and search term | Weekly | Average |
|---|---|---|
| AI brings more work, not lessworkplace burnout | 0.9 | |
| AI waste and bug-fixing burdenAI errors | 0 | |
| Longer working hours, invisible extra workovertime | 2.7 | |
| Burden shifted to managers and HRmanagerial tasks | 1 | |
| Lagging company-level productivityAI productivity | 1.9 |
World
| Topic and search term | Weekly | Average |
|---|---|---|
| AI brings more work, not lessworkplace burnout | 1.2 | |
| AI waste and bug-fixing burdenai workslop | 0 | |
| Longer working hours, invisible extra workoverwork | 1.2 | |
| Burden shifted to managers and HRmiddle managers | 1.3 | |
| Lagging company-level productivityai productivity | 16.2 |
Google Trends relative search index over the period studied, weekly average. A narrative and its sub-threads appear in a single comparison: 100 is the group's most-searched day, and the figures can only be compared within the group. The search term was chosen by machine analysis; it measures how much the term is searched, not how widespread the narrative is.
Who is affected and how
- employee
- Verifying and correcting AI output is work time rarely recognized. It is worth recording for a week how much time it takes and using this to argue for workload redistribution.
- manager
- More people are receiving managerial tasks without a pay raise, and more than half of leaders find it harder to evaluate work produced with AI. It is advisable to assign quality control as a distinct role and rewrite evaluation criteria.
- HR professional
- Burnout and turnover are hidden costs of AI. Workload and mental health should be measured among AI implementation indicators.
Hungary
Domestic footprint during the period under review is minimal. There is only one source: according to HR Portál, Hungarian HR professionals still spend half their time on administration despite AI, because expected time savings do not materialize due to poorly designed workflows.
Weekly trend
Turning points
90% of executives report no AI productivity gain, yet some firms keep cutting jobs90% of leaders have not seen productivity gains from AI, yet some companies continue to cut staff. This is the core paradox of the narrative.Shopify CEO says AI-proof-of-need rule led staff to throw "slop grenades" at each otherThe CEO of Shopify stated that employees are throwing weak AI content at each other like "slop grenades." This gave a name to the phenomenon.Regular AI users work more: mental load and time spent fixing "AI slop" riseAccording to a BCG study covering 14 markets and 11,749 workers, 41% of regular users reported increasing mental load. The burden has thus become measured data.
Related narratives in the News magazine
Supporting stories
Gartner: 95% of HR leaders use AI, but only one in five has seen significant resultsHR professionalAI training should also be planned for newly emerging roles, and its results measured.managerThe ROI of AI adoption needs to be measured, because usage does not equal impact.Gallup: workplace AI adoption up six points in Q2, but most jobs not redesignedmanagerAs AI use expands, jobs and processes must be redesigned too, or the benefits will not materialise.HR professionalWithout updated job descriptions and goals, AI use is only an added burden.employeeIt is worth documenting time savings, as they can serve as the basis for a redesigned role.Tech leaders say AI means less work, while staff report working up to 90 hours a weekemployeeDespite AI promises, expect extreme workloads at tech companies, so clarify working hours during negotiations.HR professionalBurnout risk and turnover must be managed when the AI pace involves extremely long working hours.90% of executives report no AI productivity gain, yet some firms keep cutting jobsemployeeWorkers should document overload after layoffs and report it to their manager.managerManagers should measure the real productivity effect of AI before using it to justify headcount cuts.HR professionalHR professionals should monitor the engagement and workload of staff who remain after layoffs.Glean: AI creates hidden workload for Australian employees, with 6.5 hours a week spent on oversightemployeeTell your manager how much time you spend checking AI so workloads can be planned.HR professionalTrack the AI oversight burden alongside burnout and turnover indicators.AI-driven layoffs may undermine the productivity gains companies expectmanagerBefore AI-based headcount cuts, the economics, the evidence and the reversibility should be tested separately.employeeRehiring often comes with lower pay, so it is worth thinking through your negotiating position in advance.HR professionalThe costs and loss of knowledge from rehiring after layoffs should be included when evaluating the plan.AI payoff only partly visible: McKinsey, PwC and TechCrunchmanagerReturns must be measured; costly deployment alone does not deliver results.HR professionalEmployees are using AI, but profit comes slowly; this is a planning risk.Shopify CEO says AI-proof-of-need rule led staff to throw "slop grenades" at each othermanagerQuality rules should accompany the expectation to use AI.employeeHanding over an AI draft without checking it burdens colleagues.CHROs say AI creates 'invisible' workloadHR professionalWorkload assessments should also measure the checking and correction tasks that AI generates.managerProductivity expectations should be set with the extra work surrounding AI in mind.employeeIt is worth documenting the additional tasks linked to AI.HR staff still spend half their time on administration despite AIHR professionalWorkflows must be put in order before adopting AI, otherwise administration will not fall.managerTo get a return on HR automation, invest in process design as well.AI is pushing more people into managerial work without a pay riseemployeeIf AI gives you managerial duties, raise this in role and pay negotiations.managerReassigning tasks without adjusting pay can breed dissatisfaction.HR professionalJob descriptions and pay must be aligned with the tasks that AI has changed.Skill erosion and a 'verification tax' emerge as side effects of AI at workmanagerJuniors should deliberately be given the chance to learn tasks manually, and time for checking AI output should be built into plans.early-career professionalTo gain experience, it is worth practising the basics without AI as well.employeeChecking AI output is new work that should also be reflected in performance reviews.Most UK workers use AI but save little time, Deloitte survey findsmanagerAI use alone does not deliver time savings; workflows must be redesigned too.employeeTime saved with AI is hard to measure, so apply it to specific tasks.Workers spend more time correcting AI errors than doing actual work, surveys findemployeeTrack and report the time spent checking AI output, as it determines the real productivity gain.managerInclude correction time in productivity metrics, otherwise the return on AI investment is distorted.More than half of managers find it harder to evaluate staff work produced with AImanagerRedesign how goals and output quality are defined for AI-accelerated work, rather than measuring volume.employeeDocument which parts of the work involved human judgment, so your contribution is visible at review time.HR professionalRework the evaluation system for roles that use AI.Regular AI users work more: mental load and time spent fixing "AI slop" riseemployeeAI does not automatically reduce working hours; checking outputs also takes time.managerWhen measuring productivity, also measure time spent on corrections, and set limits against pressure to speed up.HR professionalMonitoring burnout risk and workload is a new HR task.Nearly half of workers say AI worries affect their mental health, survey findsHR professionalPsychological strain should be treated as a risk of AI rollout, with support and clear communication.managerUncertainty in teams has a direct effect on engagement.
Machine analysis based on story titles and summaries, with a verification pass. The source of the facts is the article linked at each story.