The Invisible Layoff: Acquisitions and the Erosion of Labour in the Shadow of AI Hysteria

The one-line version: AI promises a transformation still taking shape; an acquisition rewrites thousands of lives overnight on a single board signature — and almost nobody is talking about the second one.

The key takeaways

Most acquisitions miss their targets — 70% to 90% of them, by Christensen’s count. When the promised revenue synergies don’t arrive, the first lever pulled is labour cost.
About a third of the jobs disappear. Swedish administrative data puts average job cuts in merged firms at 30%, with employees on both sides earning roughly 9% less than peers five years on.
The first knife goes to the back office — finance, HR, legal, IT. Biogen cut about a third of Reata’s staff two weeks after closing, citing “existing synergies” outright.
The AI picture is more nuanced. Across 41 countries, the junior share of headcount falls at AI-adopting firms — but mostly because senior hiring rises, not because junior hiring collapses.
The US layoff numbers point at AI — 101,743 attributed to it in the first half of 2026 versus 12,667 to M&A. But those are stated reasons, and nearly half the firms blaming AI had weak or declining AI usage.
AI’s effect is wildly uneven by country. Postings in the most exposed occupations fell 5.8% in high-income countries and not significantly at all in middle- and low-income ones. Acquisitions happen everywhere.
A severance package doesn’t cover the real loss — the destruction of what you built, the broken career rung, and equity options that over 70% of leavers never exercised because they couldn’t afford to.
The purest case is the one with no statistics at all: roughly 5,000 couriers in Türkiye lost 85–90% of their income without a single layoff being announced, because technically nobody was ever employed.

The modern working world is in the grip of a collective anxiety attack: is AI going to take my job? The headlines explain how algorithms will replace white-collar workers, how language models that write code in seconds will make software engineers redundant, how autonomous systems will consign call centres to history.

That dystopian projection points at a real question. But its noise covers another one: that for decades the global workforce has been ground down by a far more concrete, faster and quieter mechanism — mergers and acquisitions. The big fish swallowing the small.

AI promises a transformation spread over time, still taking shape. Holding companies, private equity funds and technology giants swallowing small and mid-sized firms is something else: a mechanical liquidation process whose workings are long since known, capable of changing thousands of lives between one night and the next morning on a single board signature.

1. Facing the data: which threat is closer?

The world of finance polishes acquisitions with words like “growth,” “market consolidation” and “economies of scale.” The operational reality draws an entirely different picture.

Most acquisitions miss their target. The analysis by Clayton Christensen and his team, published in Harvard Business Review, puts the share of acquisitions that fail to meet their objectives at 70% to 90% (HBR, 2011). When the promised revenue synergies do not arrive, the first lever pulled to rescue the balance sheet is usually the cost of labour.

Roughly a third of the jobs disappear. According to work by Jakob Beuschlein, Jósef Sigurdsson and Horng Chern Wong of RFBerlin using Swedish administrative data, merged firms cut an average of 30% of jobs. Five years on, employees of both the acquiring and the acquired company earn around 9% less than their counterparts at comparable firms. Four fifths of that loss comes from departures, the rest from the slower wage growth of those who stay (RFBerlin Research Insight). A similar finding comes from World Bank researchers: at acquired firms, employees paid above expectation are more than three times as likely to leave (World Bank blog).

Key people walk out of the door. By EY’s much-cited report, 47% of key employees leave within the first year after a transaction and 75% within the first three (Gallup). That figure counts those dismissed and those who resign together — but a resignation handed in amid that much uncertainty is often not a real choice.

The first knife goes to the back office. The units cut first under the name of “operational synergy” are usually not engineering; they are the duplicated back-office functions — finance, HR, legal and IT (The Hartford). A concrete example: just two weeks after completing its acquisition of Reata, Biogen cut roughly a third of Reata’s staff, citing “existing synergies” outright (Fierce Pharma).

On the AI front the picture is more nuanced. Bharat Chandar and Bouke Klein Teeselink of the Stanford Digital Economy Lab examined 1.25 billion job postings and 154 million employment records across 41 countries (Stanford, 2026). At firms adopting generative AI, the junior share of headcount falls — but most of that fall comes not from junior employment shrinking, but from senior employment rising. On total employment there are as-yet-unsettled findings pointing to a modest increase.

So AI, for now, is narrowing the door for new entrants; M&A is liquidating wholesale the people who have been producing value inside for years.

2. The big-fish strategy: asset hunting and the language of “synergy”

When a large technology or media giant buys an independent, innovative studio, it is usually announced as a “growth partnership.” In practice the process often turns into a three-stage cycle.

A. Eliminating duplication

The indispensable item in the acquisition decks investment bankers put before boards is “cost synergies.” The parent company already has a finance department, a legal counsel, an HR infrastructure and servers. From day one, the same units at the acquired company are labelled duplicates. That was the picture when Lloyds TSB took over HBOS too: the savings target of over £1 billion rested on merging branches and back offices (Personnel Today).

B. Hunting intellectual property and customers

Large structures often buy a company not to protect its people but to absorb its patents, its customer portfolio or its brand value. The moment the technology transfer is complete, the team that built the product from nothing becomes a load to be carried.

The games industry is the most visible laboratory of this mechanism:

Electronic Arts bought legendary studios such as Westwood Studios, Bullfrog Productions, Pandemic Studios and Visceral Games, dissolved their intellectual property into itself, and closed the studios one by one.
Microsoft, a few months after its $69 billion acquisition of Activision Blizzard, laid off 1,900 people in its gaming division in January 2024 (GameSpot). In May 2024 it closed Arkane Austin, Tango Gameworks and Alpha Dog Games — studios inside ZeniMax, which it had bought in 2021 (Windows Central). Tango, the studio it shut, had won acclaim a year earlier with Hi-Fi Rush.
Embracer Group collected dozens of independent studios on debt during the cheap-credit era. When a planned $2 billion investment deal collapsed it restructured: between July and December 2023 it laid off 1,387 people — roughly 8% of its global headcount — cancelled 29 games and closed long-established studios such as Volition (VGC). In the months that followed, the decline in headcount, together with studio sales, was measured in thousands. The message CEO Lars Wingefors gave investors at the time sums the situation up: the company’s core principle was to maximise shareholder value under any circumstances (GameSpot).

But this is not a story peculiar to the games industry:

Telecoms: while getting the Sprint merger approved by regulators, T-Mobile argued under oath before Congress that the merger would create jobs “from day one.” Two months after the merger it closed Sprint’s inside sales unit. The call announcing the closure lasted just six minutes (Android Authority). In 2023 it cut a further 5,000 people, mostly from corporate and technology roles. By GeekWire’s review, the company’s headcount was already about 9,000 below the two firms’ pre-merger combined total even before this last wave (KUOW).
Retail and private equity: Bain Capital, KKR and Vornado took over Toys “R” Us in 2005 in a leveraged buyout. The company’s debt rose from roughly 30% of its capital to 78%. In the 2018 liquidation, 33,000 employees lost their jobs. According to a letter from members of Congress, the funds extracted more than $500 million of value from the company over that period (US House of Representatives). Employees’ severance payments were cancelled. KKR, for its part, attributed the collapse to market conditions created by e-commerce (Bisnow).
Pharma and banking: as the Biogen–Reata and Lloyds–HBOS cases show, whichever sector “synergy” is written down in on paper, in practice it usually means the same thing: closing the duplicated headcount.
Japan: Toshiba was taken off the stock exchange by a $13 billion buyout from a private equity consortium led by Japan Industrial Partners. A few months later, in a country known for its lifetime-employment culture, it decided to cut up to 4,000 jobs — about 6% of its domestic workforce (Japan Times).
Türkiye: Getir was a Turkish start-up that reached a $12 billion valuation in 2022. After a battle for control with the founders, the Abu Dhabi sovereign fund Mubadala took over the company’s management and moved to sell it off piece by piece (AGBI). In February 2026, Getir’s food delivery arm was sold to Uber for $335 million (Forbes Türkiye). While the ownership fight continued in the London and Amsterdam courts, the real price was paid by the people on the ground.
Türkiye, from the ground: Uber took over 85% of Trendyol Go in 2025 and Getir’s food and grocery delivery businesses in 2026. From 9 September 2026, GetirYemek orders began to be distributed through Uber Eats–Trendyol Go, and the GetirYemek operation of Vigo, Getir’s subcontractor, came to an end (eleman.net). According to the TEHİS union, that step affects roughly 5,000 motorcycle couriers. The couriers were given no job guarantee on the new platform; the GetirYemek orders that made up 85–90% of their income were taken away from them. The union claims those orders were shifted into a system paying 20–30% less per package (TEHİS). No official redundancy announcement was made; because the couriers work as independent contractors, technically nobody was “fired.” For that reason they will never appear in any layoff statistic either. This is perhaps the purest example of the “invisible layoff” in this piece’s title. (The figure of 5,000 belongs to the union; no official number confirmed by Uber or Getir has been published.)
Türkiye, e-commerce: eBay bought a majority stake in GittiGidiyor, one of Türkiye’s first internet start-ups, in 2011, and took full ownership in 2016. In June 2022 it closed the platform and withdrew from Türkiye, citing “competitive dynamics in the market” (NTV). The closure statement contained not one line about employees; it did, however, make a particular point of noting that the closure would not affect eBay’s quarterly results (Habertürk). A twenty-year-old company was too small an item to be even a footnote on the acquirer’s balance sheet.

C. Unlocking the golden handcuffs and the founders’ exodus

In acquisitions, key founders and executives are usually given two- to four-year earnout or retention packages. Throughout that period, the shield protecting the team against the bureaucracy of the larger structure is the founding team. The day the shares turn into cash, the founders leave. A unit with no advocate left at the table becomes the first target to be written off in the holding company’s next budget squeeze.

3. The destruction a severance package cannot hide

The classic defence from those who look at this from a strict free-market perspective runs: “They get their statutory severance anyway, sometimes even six- to twelve-month packages. What is the problem?” That approach sees a person as a unit of consumption with money landing in their account at month’s end. But the loss is not confined to salary.

The destruction of what was built. For engineers, designers and product managers, the meaning of the work depends on the survival of what they produce. A product developed over years of sleepless nights being thrown away by a single decision — because it “shouldn’t compete with my main product,” or because “we’ve changed our strategic focus” — leaves a question in the employee’s mind that does not close easily: what was all that effort for? After Tango Gameworks was shut, the head of its sibling studio Arkane Lyon, Dinga Bakaba, calling the decision plainly “horrible” was a rare expression of that feeling from inside the industry (Game World Observer).

The breaking of career momentum. According to the RFBerlin study cited above, those who lose their jobs either fall into unemployment or are forced to move to smaller, lower-paying firms. The worst affected are older workers and those with relatively less education (RFBerlin). A severance package rescues a few months; it does not rescue a broken rung on a career ladder.

The equity promise coming to nothing. In the start-up world, the thing expected to compensate for a low salary is stock options. According to Carta’s data, more than 70% of vested options in 2025 went unexercised as the employee left the company. The main reason is that most employees do not have the cash required to exercise them (Carta). For someone who has been laid off, that decision is usually squeezed into a 90-day window. And when an acquisition happens, the money is distributed first to liquidation preferences and debts. Within that structure, the ones who get a cash exit are mostly the investors and the senior management; the junior and middle ranks who carried the company on their backs either get no share at all or get very little.

For someone who has lost their job, the matter is not only next month’s rent. Being part of a team, contributing to a vision and producing something is erased with a single line in a financial engineering spreadsheet.

4. Why is AI talked about so much?

There are two reasons for this, and both of them are real.

First: the new kid in town. AI is like the “cool” kid who has just arrived at school. Everyone is talking about them; they lead every conference, every investor deck, every headline. The narrative of “machines surpassing human intelligence” generates clicks like a science fiction film. Mergers, meanwhile, are the student who has sat in the same classroom for years and whom nobody notices.

Second: AI really is transforming things. One company buying another is a strategic or financial decision. It is taken once and affects one company. AI, whether we want it or not, is a general-purpose technology changing how work is done at every company. Like the spreadsheet or email, it is spreading to every desk; it is transforming some sectors and destroying some jobs. So it being talked about more is, in a way, natural.

But something is getting lost in the shadow of that second reason. Having “been around for ages” does not make a risk harmless; it only makes it invisible.

5. From the employee’s view: which is the more concrete risk?

The real question is this: as an employee, which is more likely to put me out of a job? My company connecting to an AI API and finding me redundant — or my company being swallowed by a bigger fish and my position, counted as a duplicate, being closed?

At first glance the numbers point at AI. According to Challenger, Gray & Christmas, which tracks US layoffs by stated reason, in the first six months of 2026 companies attributed 101,743 layoffs to AI and only 12,667 to mergers and acquisitions (Challenger Report, June 2026). The picture was similar across the whole of 2025: 54,836 against 25,622 (Challenger Report, February 2026). It would not be honest to ignore this data: AI is now the reason companies state most often for layoffs.

But this data covers only the US, and the world is not only the US. AI’s effect on the labour force differs enormously from country to country. According to a World Bank researchers’ study examining 555 million job postings across 84 countries, postings in the occupations most open to AI substitution fell by an average of 5.8% in high-income countries. In middle- and low-income countries the effect remained very small and was not statistically significant (AMRO seminar, World Bank study). According to the ILO, one in every four jobs in the world is exposed to generative AI to some degree. But the share of jobs in the highest-risk group is only 3.3%, and — as the ILO itself stresses — exposure does not mean job loss (ILO; The People Space summary). So for someone working in Senegal, Bangladesh or Türkiye, AI’s risk of directly costing them their job today is far lower than it is for a software engineer in the US. Acquisitions, on the other hand, happen everywhere — in Tokyo and in Istanbul alike, as we saw above.

But these figures fail to show the whole picture for three important reasons:

These are stated reasons. Challenger records the reason the company gives. “We are transforming with AI” sounds a great deal better to an investor than “we over-hired.” In Revelio Labs’ analysis, nearly half of the companies attributing layoffs to AI had weak or declining AI usage (AI Understanding). At some companies, “AI” is new packaging on an old-fashioned cost cut.
Acquisition cuts are spread out and relabelled. An acquisition does not announce its cuts in one go; it spreads them over years and calls them “restructuring,” “cost optimisation” or “closure.” T-Mobile described a wave of cuts two years after the merger as “ordinary organisational shifts” (Fierce Network). The 33,000 employees of Toys “R” Us were likewise written into the statistics of a “bankruptcy,” not an acquisition.
Individual risk is different from the aggregate. What matters is what happens when it happens to you. If your company is acquired, the Swedish data says an average of 30% of jobs in the merged entity disappear. If your company adopts AI, Stanford’s 41-country data shows no marked fall in total employment; what changes is the composition of the headcount.

And some of the jobs counted as “fired because of AI” come back. Sweden’s Klarna announced with great pride that its AI assistant was doing the work of 700 customer service agents. A few months later CEO Sebastian Siemiatkowski admitted that this cost-driven move had lowered service quality, and the company began hiring humans again (Entrepreneur). The “catch the hype, fire, then take them back” cycle is still painful for the employee — but only the firing part shows up in the statistics.

	Acquisition	AI
How it arrives	On a single signature, sudden	Spread over time, gradual
Who it hits first	The duplicated back office; senior and expensive staff	Entry level; tasks open to automation
How it is labelled	“Synergy,” “restructuring”	“Transformation,” “efficiency”
How much it is discussed	Almost never	Constantly

In short: AI is a spreading, growing risk that deserves attention. An acquisition is a risk that is sharper when it happens to you, more measurable, and talked about far less. An employee should ask themselves two questions: “How much of my job can be automated?” and — at least as important — “Can my company be sold?”

6. Conclusion: why do we talk about AI and not M&A?

The fear of AI is new, visible and fiercely debated; that debate is necessary. Mergers, the formation of holding companies and private equity funds breaking firms up and selling them, on the other hand, are considered too bureaucratic, too colourless and too “ordinary.”

What is more, the victims of these processes often cannot speak. Exit packages frequently come with confidentiality and non-disparagement clauses; taking the package means agreeing to stay silent about what happened. That silence makes M&A-driven unemployment one of the largest but least discussed labour tragedies.

That AI raises productivity and transforms some professions is a fact; following that transformation is necessary too. But the thing that takes an employee’s job, identity and space of production away from them overnight is, today, usually not an algorithm that has gained consciousness. It is the corporate consolidations that reduce a human life to a line reading “cost synergy” in order to tidy up a balance sheet. We need to start ranking our fears according to the real risks.

Sources

Academic studies and reports

Beuschlein, J., Sigurdsson, J. & Wong, H. C. (2025). The Labor Market Consequences of Acquisitions. RFBerlin Discussion Paper. Summary: Who Loses When Firms Are Bought.
Chandar, B. & Klein Teeselink, B. (2026). How Does AI Change Labor Demand? Evidence from 41 Countries. Stanford Digital Economy Lab.
Christensen, C. M., Alton, R., Rising, C. & Waldeck, A. (2011). The Big Idea: The New M&A Playbook. Harvard Business Review.
World Bank. Firm consolidations can have long-lasting impacts on workers’ labor market outcomes. Let’s Talk Development blog.

The Invisible Layoff