AI Job Cuts 2026: What 99% of CEOs Are Planning
The Mercer Global Talent Trends 2026 report makes one finding impossible to ignore: virtually every CEO surveyed expects to reduce headcount because of AI within two years. The question is no longer whether this will happen, but how companies manage the transition.
The Mercer Global Talent Trends 2026 report surveyed nearly 12,000 executives across 16 countries and 16 industries. Its headline finding: 99 percent of CEOs plan AI-driven workforce reductions within two years. Two thirds target cuts of 1 to 10 percent; nearly one third plan reductions of 11 to 20 percent. In parallel, employee wellbeing has collapsed: only 44 percent of employees are thriving, down from 66 percent in 2024. In Germany, 19 percent of AI-using companies have already eliminated positions according to Bitkom 2026, even as 67 percent of German firms expect no net negative impact on employment. The entry-level paradox compounds the risk: 43 percent of CEOs plan to cut junior roles, undermining the very talent pipeline that will manage tomorrow's AI-augmented organisations. Companies that treat this as a pure cost exercise, rather than a transformation challenge, will pay for that miscalculation in future capability gaps.
What Is Happening: The Mercer Numbers
Almost all CEOs worldwide expect to lay off employees due to AI within the next two years. The Mercer Global Talent Trends 2026 report surveyed nearly 12,000 executives and professionals across 16 countries and 16 industries, making it one of the most comprehensive readings of executive intent on workforce transformation currently available. The headline number is stark: 99 percent of surveyed CEOs expect to reduce headcount as a direct consequence of AI deployment.
The scale of planned reductions is significant. This is not a wave of single-digit adjustments affecting only the most easily automated roles. A substantial share of CEOs are planning structural workforce changes that will reshape the composition of their organisations within a short planning window.
The 99 percent figure reflects a near-universal executive consensus. What varies is the magnitude and the approach. Some organisations are already executing workforce reductions; others are in planning phases. The Mercer data does not predict uniform disruption across all sectors and job families, but it does confirm that AI-driven headcount changes are no longer a fringe scenario discussed by technologists. They are a mainstream management priority being planned at the highest level of global organisations.
The Mercer findings also reveal that 65 percent of executives expect to redeploy or reskill between 11 and 30 percent of their workforce. This signals that the most forward-looking organisations are not treating this purely as a headcount reduction exercise. They are attempting to shift the composition of their workforce rather than simply reduce it. The gap between organisations that invest in reskilling and those that simply cut will be visible within the next three years in capability, morale and competitive positioning.
The Mercer 2026 data represents the most comprehensive survey of CEO intent on AI-driven workforce changes to date. The consensus is near-total. European companies that are not yet running a parallel workforce planning process alongside their AI deployment programme are behind where their global peers are already operating.
Which Jobs and Sectors Are Hit First
Job postings are already declining, and they function as an early indicator of where AI displacement is materialising ahead of formal announcements. The data between Q3 2024 and Q3 2025 shows concentrated declines in roles characterised by structured communication, information coordination and repetitive knowledge work.
The steepest decline is in call center roles, down 67 percent in job postings over twelve months. This is not coincidental. The combination of large language model-powered conversational agents and workflow automation has made scalable AI handling of tier-one customer service enquiries technically and economically viable for organisations at scale. Allianz Partners has already announced the replacement of 1,500 to 1,800 call center workers with AI systems, making it one of the most visible real-world implementations of this transition in the European insurance sector.
Copywriting positions have fallen by 53 percent over the same period, reflecting the direct impact of generative AI on text production workflows. Project management postings are down 48 percent, driven by AI tools that automate status tracking, resource scheduling, meeting summaries and dependency mapping. IT consulting postings have declined by 34 percent, a development that reflects the ability of AI coding assistants and automated architecture review tools to compress work that previously required sustained specialist engagement.
Lufthansa has announced the reduction of 4,000 administrative roles by 2030 as part of its AI-driven operational efficiency programme. These are not warehouse or manufacturing roles. They are white-collar, knowledge-intensive positions in areas like documentation, scheduling and internal reporting, exactly the profile of roles that appear across every sector where AI adoption is accelerating.
The entry-level concentration of planned cuts is structurally significant. Mercer data shows that 43 percent of CEOs specifically plan to eliminate junior roles, more than double the 17 percent who indicated the same intention in 2024. Junior positions are the entry point through which the next generation of professionals builds domain judgment, institutional knowledge and the contextual understanding of how organisations actually operate beneath their formal procedures. Organisations that hollow out this layer will face a capability deficit in their mid-senior ranks three to five years from now that no amount of reskilling budget can fully compensate for. The AI Knowledge Atrophy Paradox describes exactly this dynamic in detail.
Sector and role data confirm what the CEO survey signals: the displacement is concentrated, not uniform. Call centers, copywriting, project management and IT consulting are leading indicators. The disproportionate targeting of junior roles is the most strategically consequential trend because its negative effects will not be felt immediately, but will compound over years.
European Perspective: Between CEO Consensus and Reality
In Germany, a striking contrast emerges: globally, nearly all CEOs announce headcount reductions driven by AI, but 67 percent of German companies expect no negative net impact on total employment according to Bitkom 2026. Understanding this gap is essential for European decision-makers who risk either dismissing global signals as irrelevant to their context or importing US-framed narratives without accounting for the structural differences in European labour markets.
German AI adoption has undergone a significant shift. Bitkom 2026 reports that 41 percent of German companies now actively use AI, up from 17 percent a year earlier. This adoption rate has a direct employment dimension: 19 percent of these AI-using firms have already eliminated positions. Given that over 40 percent of the German economy now uses AI in some form, the 19 percent figure represents a meaningful and growing portion of the workforce already affected.
The long-term projection of 1.6 million affected jobs over 15 years comes with an important caveat: most analyses, including those from the IAB, expect this to be partially or fully offset by new job creation in AI operations, oversight, governance and adjacent technical roles. The net employment picture is genuinely uncertain. What is not uncertain is that the composition of the German workforce will change substantially, and that this change will not be evenly distributed across sectors, regions or skill profiles.
The Siemens model is worth examining as a European alternative to simple headcount reduction. Rather than eliminating 6,000 positions outright, Siemens committed to retraining those employees for AI-adjacent roles. This approach is consistent with the co-determination culture embedded in German corporate governance, but it requires early planning, significant investment in learning infrastructure and leadership willingness to treat workforce transformation as a multi-year programme rather than a quarterly cost adjustment.
The IAB 2026 data introduces a regional dimension that national aggregate figures obscure. Employment decline is expected in 10 of 16 German federal states, reflecting the fact that AI-driven displacement does not hit all economic geographies equally. States with higher concentrations of routine administrative and service work face greater near-term exposure than those with stronger research, engineering and high-complexity service sectors.
The German market is not immune to global CEO workforce plans, but its structural context, co-determination, collective bargaining, and an established culture of reskilling, creates more space to manage the transition thoughtfully. The 19 percent of AI-using firms that have already cut jobs are a leading indicator, not an outlier.
The Wellbeing Problem
Parallel to announced layoffs, employee wellbeing is collapsing, and this is happening regardless of whether actual layoffs have occurred at any given organisation. The Mercer data reveals that the mere anticipation of AI-driven changes is producing measurable deterioration in the psychological and emotional conditions that drive sustained organisational performance.
The decline in thriving employees from 66 percent to 44 percent in two years is a historic drop for a metric that typically moves slowly. When fewer than half of your workforce considers itself to be thriving, the consequences extend well beyond individual wellbeing. Engagement, retention, risk-taking, collaboration and the discretionary effort that drives innovation all decline alongside it. This is not a soft metric. It has direct operational consequences.
The fear of AI job displacement has risen from 28 percent to 40 percent of employees over the same period. This increase precedes, rather than follows, actual job cuts at most organisations. Employees are responding to signals in the public environment, in their organisation's communications and in the arrival of AI tools in their daily workflows. The anxiety is real even where the restructuring has not yet occurred.
The 62 percent figure who say leadership underestimates the emotional impact is particularly telling when set against the 19 percent of HR professionals who consider monitoring these impacts important. There is a large and consequential gap between how employees experience AI-related change and how organisations are equipped to detect and respond to it. Most companies are flying blind on the human cost of their transformation programmes.
This is not an argument against AI adoption. It is an argument for treating the human dimension of that adoption with the same rigour applied to technical implementation. The organisations in the Mercer data that report higher wellbeing scores consistently show stronger AI adoption outcomes, better change acceptance and lower voluntary attrition during restructuring periods. Wellbeing is not a secondary consideration to workforce transformation. It is a primary enabler of it.
The collapse in employee thriving rates is a leading indicator of organisational fragility. Companies that announce AI-driven changes without investing proportionally in communication, support and psychological safety will experience higher attrition, lower productivity and weaker transformation outcomes precisely when they can least afford it.
Challenges and Risks
The biggest blind spot in current AI workforce planning: companies are cutting exactly the positions where talent develops the experience and judgment that makes organisations resilient. The concentration of planned cuts in junior roles and high-volume knowledge work functions is removing the system's capacity to self-renew.
65 percent of executives expect to redeploy or reskill 11 to 30 percent of their workforce, but only 32 percent believe their teams currently have the capacity to collaborate effectively with AI systems. The gap between the ambition and the capability is the primary execution risk for AI workforce transformation programmes.
AI as a restructuring cover: A non-trivial share of announced AI-driven layoffs conflates genuine automation-driven efficiency with strategic restructuring that would have occurred regardless. When AI becomes the stated rationale for decisions driven by competitive pressure, post-merger integration or financial underperformance, it distorts the public data on AI displacement and, more importantly, erodes internal trust when employees eventually recognise the framing as incomplete. Organisations that use AI as a shield for other decisions pay a credibility cost that compounds over subsequent change programmes.
Cost overruns are common: 33 percent of German AI users report that AI has cost more than expected. This figure from Bitkom 2026 reflects a pattern visible across markets: the direct procurement cost of AI tools is typically lower than the total cost of deployment, integration, data preparation, employee training, governance overhead and ongoing model maintenance. Companies that model AI ROI on tool cost alone systematically underestimate the investment required and overestimate the speed of returns.
Redeployment complexity: 65 percent of executives expect to redeploy or reskill between 11 and 30 percent of their workforce. This is a substantially more complex undertaking than headcount reduction. It requires mapping current competencies against future role requirements, designing learning pathways, managing psychological transitions for employees who are moving between functions, and building line management capability to integrate newly reskilled staff into changed workflows. Most organisations have limited prior experience with reskilling at this scale and timeline.
The collaboration capability gap: Only 32 percent of executives believe their teams can currently collaborate effectively with AI systems. If AI is to deliver the productivity gains that justify planned headcount reductions, the remaining 68 percent of organisations face a significant upskilling challenge before realising those gains. The sequence matters: organisations that reduce headcount before building AI collaboration capability risk a gap period where they have fewer people and less productivity.
EU AI Act compliance deadline approaching: AI systems that influence or inform personnel decisions, including hiring, promotion, performance assessment and redundancy selection, are classified as high-risk under Annex III of the EU AI Act. Obligations for these systems, including human oversight requirements, transparency documentation, conformity assessments and registration, apply from December 2027. Organisations that have not yet audited which AI tools are in scope for Annex III risk non-compliance in the most reputationally sensitive category of AI use. For full detail, see the EU AI Act High-Risk Deadlines analysis.
What Companies Should Do Now
Companies planning AI-driven workforce changes should treat this as a transformation project, not a pure cost-cutting exercise. The Siemens model demonstrates that reskilling alongside right-sizing is achievable with early planning and genuine investment in workforce capability. The organisations that will perform best through this period are those that are deliberate about what they are preserving, not just what they are reducing.
- Conduct a competency audit before making any cuts. Map current role profiles against near-term automation likelihood and longer-term strategic value. Distinguish between tasks that AI will eliminate, tasks that AI will change and tasks that require sustained human judgment regardless of AI capability. This audit is the foundation for every subsequent decision. Without it, workforce reduction and reskilling programmes are operating on assumptions rather than evidence. Connect to your Convergence Era Strategy to ensure workforce planning aligns with the broader transformation roadmap.
- Prioritise reskilling as a competitive signal, not just an ethical choice. Mercer data shows that 77 percent of investors view workforce reskilling programmes favourably as an indicator of long-term competitive positioning. Companies that publicly commit to reskilling rather than pure headcount reduction attract both talent and capital more effectively. This is not altruism. It is a reputational and financial advantage that compounds over time.
- Invest in transparent, sustained communication. 62 percent of employees say leadership underestimates the emotional and psychological impact of AI-driven change. Transparent communication does not mean announcing every detail of plans that are still forming. It means acknowledging the uncertainty honestly, explaining the principles guiding decisions, creating channels for questions and committing to updates on a defined schedule. The organisations with the best outcomes from AI transformation communicate more, not less, in the ambiguous middle period.
- Protect entry-level positions deliberately. The disproportionate targeting of junior roles, planned by 43 percent of CEOs, is the most structurally damaging pattern in the Mercer data. Decision-makers who look only at the immediate cost calculus are missing the five-year consequence for their talent pipeline. Organisations that preserve structured entry-level pathways alongside AI deployment build the next generation of professionals who understand both the human work and the AI systems. Those that do not will face a capability shortage they cannot hire their way out of. The AI Knowledge Atrophy Paradox details exactly how this degradation compounds.
- Start EU AI Act compliance preparation now for HR and personnel AI systems. AI tools that influence hiring, performance management, promotion or redundancy decisions are classified as high-risk under EU AI Act Annex III. Compliance obligations including human oversight requirements, conformity assessments and registration apply from December 2027. Organisations need approximately 12 to 18 months of preparation for Annex III compliance in typical enterprise environments. Starting in mid-2026 provides adequate runway. Starting after that creates acute time pressure on legal, HR and IT teams simultaneously. The full deadline framework is covered in the EU AI Act High-Risk Deadlines analysis. The broader context of why German companies show a different pattern is explored in the German Mittelstand AI Boom article.
The companies that will manage AI-driven workforce transitions most successfully are those that treat the human dimension with the same seriousness they apply to the technical and financial dimensions. That means competency auditing before cutting, reskilling investment alongside reduction, and genuine communication throughout. The Siemens model exists because early planning created the option to execute it. Late planners do not have that option.
Further Reading
Frequently Asked Questions
The Mercer Global Talent Trends 2026 report surveyed nearly 12,000 executives and professionals across 16 countries and 16 industries. It found that 99 percent of CEOs expect to reduce headcount due to AI within the next two years. Of these, 67 percent plan reductions of 1 to 10 percent of their workforce, while 32 percent anticipate cuts of 11 to 20 percent. Additionally, 43 percent specifically plan to eliminate junior positions, up from 17 percent in 2024.
According to a long-term projection, approximately 1.6 million jobs in Germany could be affected by AI automation over the next 15 years. However, experts expect this to be balanced by new job creation in AI-adjacent roles. In the near term, Bitkom data from 2026 shows that 19 percent of German companies actively using AI have already cut positions. The IAB (German Institute for Employment Research) expects employment decline in 10 of 16 German federal states, though definitions and timelines vary significantly across forecasts.
Job posting data between Q3 2024 and Q3 2025 shows the steepest declines in call center roles (down 67 percent), copywriting (down 53 percent), project management (down 48 percent) and IT consulting (down 34 percent). These declines reflect early displacement of tasks involving structured communication, repetitive information processing and coordination work. Real-world examples include Allianz Partners replacing 1,500 to 1,800 call center workers with AI systems and Lufthansa planning 4,000 administrative role reductions by 2030.
The entry-level job paradox refers to the pattern where companies disproportionately target junior positions for AI-driven cuts. Mercer data shows 43 percent of CEOs plan to eliminate junior roles, more than double the 17 percent who said the same in 2024. The paradox is structural: these are exactly the positions where the next generation of professionals builds the judgment, domain knowledge and institutional understanding that becomes senior expertise. Cutting entry-level roles today undermines the pipeline of experienced talent for tomorrow.
European companies should approach AI-driven workforce changes as transformation projects rather than cost-cutting exercises. Key steps include: conducting a competency audit to distinguish tasks automatable in the near term from those requiring sustained human judgment; prioritising reskilling, which carries 77 percent investor approval as a competitive advantage signal; communicating transparently, as 62 percent of employees say leadership underestimates the emotional and psychological impact; protecting entry-level positions where future senior talent develops; and preparing for EU AI Act compliance, since AI systems that influence personnel decisions are classified as high-risk under Annex III with obligations starting from December 2027.
Globally, 99 percent of CEOs announce AI-driven headcount reductions, but 67 percent of German companies expect no negative net impact on total employment, according to Bitkom 2026. This gap reflects several factors: Germany's co-determination model gives employee representatives formal influence over restructuring decisions; collective bargaining agreements constrain rapid workforce reduction; and German companies often emphasise reskilling and redeployment as a first response. The Siemens example is instructive: rather than cutting 6,000 positions outright, the company retrained those employees for AI-adjacent roles. This approach is more common in Germany than in North America or the UK.