The world's most profitable social media company started notifying employees of mass redundancies in its Asian hub before most of the city-state had woken up
At 4 o'clock on a Wednesday morning in Singapore, Meta Platforms began sending termination notices to thousands of employees worldwide. The city-state, home to one of the company's major regional hubs, was chosen as the first to receive word - a logistical decision reflecting time zones, but one that left employees waking to the news that their roles no longer existed.
By the time workers in Europe and the United States log on, the scale of what is happening will become clear. Meta is cutting approximately 8,000 positions globally - roughly 10 per cent of its 78,865-person workforce - in a restructuring that its Chief People Officer Janelle Gale described in an internal memo as designed to help teams "operate with a flatter structure with smaller teams of pods and cohorts that can move faster and with more ownership."
Simultaneously, the company is reassigning 7,000 employees to newly created AI-focused teams, cancelling 6,000 open roles it had planned to fill, and signalling that further cuts will follow in the second half of the year. The total effective headcount impact - roles eliminated plus positions never created - is 14,000. Additional rounds are planned for August and the northern autumn.
This is the third wave of Meta layoffs in 2026 alone. It is the largest single corporate restructuring event in Asia's technology sector this year.
The business logic - and why it does not explain the human reality
Meta is not cutting because it is struggling. The company reported revenue of $201 billion in 2025, up 22 per cent year on year, with fourth-quarter net income of $22.8 billion. Its stock has risen sharply this year. The cuts are happening because the company has committed $115 billion to $145 billion in capital expenditure in 2026 - the vast majority directed at AI infrastructure - and has concluded that it can sustain and grow its operations with a smaller, more AI-augmented workforce.
Analysts at Evercore estimate the redundancies will generate approximately $3 billion in annual savings. Against projected AI spending that could reach $145 billion this year, that is a rounding error. The cuts are not primarily a cost measure. They are a structural statement: Meta believes that flatter, AI-native teams organised around "pods and cohorts" will outperform the layered management structure that the company has operated with for the past decade.
Mark Zuckerberg has been direct about this. He has publicly encouraged engineers to use AI agents to assist with coding. He has outlined plans to track employee devices - gathering keystrokes, mouse movements and screen content - to train AI systems. He has spent time building his own AI-powered personal assistant to handle some of his CEO duties. The direction of travel is unmistakable: human roles are being repositioned as inputs to AI capability, not as the primary source of organisational output.
The employees at the receiving end of Tuesday's notifications experienced something different. More than a thousand Meta staff had already signed a petition demanding the company refrain from collecting their data from devices in the effort to train AI. Others took to social media to post about how the threat of layoffs had affected their work and morale. HRD Asia's coverage of the employee protest earlier this week captured the dynamic precisely: workers were not resisting AI as a technology so much as resisting the inference - reinforced by the company's own public statements - that their day-to-day work was being harvested to build the systems that would make their roles redundant.
Singapore first: what it means to be an Asian hub in the AI restructuring era
The decision to begin notifications in Singapore carries significance beyond the logistics of time zones.
Singapore has positioned itself as one of Asia's most important technology and regional headquarters locations, hosting the APAC operations of dozens of the world's largest technology companies. For years, that positioning has been a source of economic strength - a pipeline of high-quality, well-compensated knowledge work flowing into the local economy.
What Tuesday's 4am notifications illustrate is that hub status does not confer protection. When a global technology company restructures around an AI-native operating model, the geographic origin of a role matters far less than its function. Coordination roles, management layers, recruiting operations, and sales support - the functions that populate regional hubs - are precisely the categories that AI-driven restructuring eliminates first.
Singapore's Ministry of Manpower's inaugural AI adoption report found that 71.5 per cent of firms in the city-state have yet to adopt AI, with only 6.2 per cent of firms reporting reduced headcount after adopting AI so far. Those figures describe local firms. The behaviour of global technology multinationals - of which Meta is the most visible current example - is running well ahead of that domestic baseline. Singapore-based employees of global technology companies are experiencing an AI restructuring wave that the city-state's own businesses have barely begun.
HRD Asia has reported on Singapore's efforts to get ahead of this dynamic, including an Economic Strategy Review committee exploring whether to require earlier retrenchment notifications so that affected workers can access support sooner. Under current rules, companies with at least 10 employees must notify the Ministry of Manpower within five working days of informing an employee of retrenchment. The question of whether that window is adequate for a company notifying thousands of employees globally at 4am is one that Singapore's tripartite partners are actively examining.
The severance picture and what HR should benchmark against
US-based Meta employees affected by the May round will receive 16 weeks of base pay plus two additional weeks per year of service, and 18 months of health coverage. The company has committed to providing immigration support for employees on work visas, a significant concern given that H-1B filing fees have risen sharply and the window for finding a new sponsor within 60 days is considerably narrower than it was during previous layoff cycles.
For employees in Singapore and across the Asia-Pacific region, severance entitlements are determined by local employment legislation rather than the US package. Singapore's Employment Act requires a minimum of one week's salary per year of service for employees with fewer than five years' tenure, rising to two weeks for those with five to ten years, and three weeks for ten or more years of service. For many employees at a company like Meta, whose typical compensation structures are weighted toward equity and bonus, the statutory minimum is a poor reflection of the economic disruption a redundancy represents.
HR leaders at organisations across the region who are watching Meta's approach as a benchmark should consider whether their own redundancy frameworks - not just their legal minimums, but their voluntary packages, outplacement support, and communication protocols - are adequate for the scale and speed at which AI-driven restructuring is now proceeding.
The legal exposure that HR must understand
The method of Meta's notifications - batched emails sent at 4am local time, with employees encouraged to work from home while the process completes - raises questions that HRD Asia has specifically addressed in its coverage of AI-driven dismissals under Singapore's emerging fairness framework.
Zhao Yang Ng, principal in the employment practice group at Baker McKenzie Wong & Leow, has been explicit: employers using AI tools in performance evaluation or workforce selection have an obligation to inform employees, to be able to explain and justify the decision-making process, and to provide meaningful appeal mechanisms. "If you cannot explain how the tool made its decision, or if it was trained on flawed data, you are exposed to discrimination claims," Ng told HRD.
Multiple Meta employees reported receiving unexpectedly negative performance ratings from managers they believed were satisfied with their work, only to learn that the ratings were being used to justify inclusion in layoff lists - a pattern described as "stack ranking by stealth" that allows companies to frame performance-based terminations as distinct from layoffs, potentially reducing severance obligations. The Workplace Fairness Act, due to come into force in Singapore by 2026 or 2027, will create new obligations in precisely this area. HR leaders whose organisations are planning AI-influenced workforce reductions need to audit their processes against that framework now.
The Singapore National Employers Federation has warned against "overly prescriptive" regulation in this area, arguing that flexibility is required for companies to restructure responsibly. That framing reflects genuine tensions in Singapore's approach: the government wants to encourage AI adoption and corporate investment while protecting workers from its most disruptive consequences. For HR professionals, the practical implication is that the regulatory direction is toward greater accountability, not less - and that building fair, transparent, and documentable processes now is both the ethically correct approach and the legally prudent one.
The boomerang risk that nobody is talking about
There is a dimension of Meta's restructuring that its investor presentations do not address, and that HR professionals should name.
Gartner now forecasts that by 2027, 50 per cent of companies that attributed customer service headcount reductions to AI will rehire staff to perform similar functions, even if the roles return under different titles. Forrester's 2026 future of work outlook reaches a similar conclusion, expecting half of AI-attributed layoffs to be quietly reversed, with jobs returning offshore or at lower wages.
Meta is not cutting roles that are genuinely redundant. It is cutting roles that it believes AI can replicate, on a timeline and at a productivity level that is projected but not yet proven. If those projections turn out to be wrong - as they have been for a significant proportion of companies that have already tried this - the knowledge, institutional memory, and human capability that was eliminated on Tuesday morning does not simply return when the company decides it needs it again.
HRD Asia has reported on the boomerang trend in detail, finding that the "layoff boomerang" - companies terminating employees only to rehire them or equivalent workers within months - is already visible in the data. For Singapore's technology labour market, that pattern creates a specific problem: the talent displaced by Meta and companies like it does not sit still. It finds roles at competitors, returns to home countries, or leaves the technology sector. When companies discover they need those skills again, they are not available at the same price or in the same location.
What HR leaders in the region should do now
The Meta restructuring is not an isolated event. The tech sector has shed more than 95,000 jobs across 247 layoff events in 2026, an average of 882 per day. Amazon cut tens of thousands of roles in January. Oracle eliminated up to 30,000. Atlassian cut 1,600. Every major company cites AI restructuring as the primary driver. The direction is consistent: traditional roles out, AI roles in, with the savings redirected into infrastructure.
For HR leaders at organisations in the region - whether at multinationals with Singapore hubs, local technology companies, or businesses in adjacent sectors watching this wave approach - several practical obligations follow.
On notification process: the 4am email is technically compliant with many jurisdictions' minimum requirements. It is not adequate human practice. Employees who are told they no longer have a job by an automated notification before they have had their morning coffee are not being treated with the dignity that Singapore's Tripartite Guidelines on Fair Employment Practices require. HR leaders who are designing or approving mass notification processes should insist on something better.
On legal exposure: if AI tools are being used at any point in the workforce selection process - whether in performance management, productivity monitoring, or role elimination criteria - the obligation to explain, document and defend those decisions is already present in Singapore law, and is about to become more demanding. HRD Asia's legal guidance on this is the starting point.
On talent retention amid the wave: hiring sentiment in Singapore has rebounded in 2026, with the government planning to upskill 100,000 employees in AI under a new national programme. The employees currently being displaced by Meta are experienced technology professionals. The organisations that move quickly to offer them meaningful roles - and the genuine capability development that Meta promised but did not deliver - will be the beneficiaries of the restructuring wave, not simply its bystanders.
On the survivor population: the 78,000-odd Meta employees who are not being terminated on 19 May are watching closely. Research consistently finds that survivor syndrome - the guilt, anxiety and reduced engagement of employees who remain after mass redundancies - produces measurable productivity losses. Combined with ongoing surveillance anxiety from the mouse-tracking controversy and the knowledge that further rounds are planned for later in the year, the conditions at Meta are as close to a controlled experiment in workforce psychological harm as corporate life produces. HR leaders at other organisations should note what it looks like from the outside, and ask whether their own AI restructuring communications are creating a similar experience internally.
Meta spent more than $100 billion building the AI infrastructure that is being used to justify removing 10 per cent of its human workforce. The employees receiving their notifications at 4am in Singapore this morning did not choose to be the test case for that theory. HR's responsibility - here, as everywhere - is to ensure that when change of this scale is imposed on people, it is done with fairness, transparency, and genuine regard for the human beings on the receiving end.
That is not a soft imperative. In Singapore, increasingly, it is a legal one.
