Did AI Really Take These Jobs? The Truth Behind the Oracle, TCS, Meta and Microsoft Layoffs

Why "Did AI Take the Jobs" Is the Most Searched Question Right Now

In short: AI is a real factor behind 2025 and 2026 layoffs, but it is rarely the only one, and companies now find it a convenient label for cuts that are also about interest rates, past over-hiring, weak earnings and plain cost-cutting. Independent tracking firm Challenger, Gray and Christmas found that AI was explicitly cited in roughly 55,000 of the more than one million job cuts announced in the United States in 2025, a real number, but a small share of the total. The sections below walk through what actually happened at Oracle, TCS, Meta and Microsoft, and show how to separate the AI story from the rest of it.

The search suggestions in the screenshots people keep sharing tell their own story. People are not just asking whether AI will take jobs some day. They are asking about software engineers, doctors, teachers, lawyers, pilots and data analysts, specific, skilled, once-secure professions. That shift in the question itself, from a hypothetical future to a list of real occupations, reflects what has actually been happening on the ground since the middle of 2025.

Four stories drove this anxiety more than any others. A viral post about 80 people losing their jobs on a five-minute call. An early-morning email from Oracle that read like a form letter. A record-breaking round of cuts at India's biggest IT employer. And a steady drumbeat of thousands more at Meta and Microsoft, two companies that are simultaneously spending more than ever on AI while also spending less on people. None of these stories exists in isolation, and understanding how they connect is the point of this article.

The 5-Minute Call: What Happened to an 80-Person Support Team

In September 2026, a Reddit post titled "Laid off over a Google Meet" spread quickly across Indian social media and then into national news coverage. The person who wrote it said they worked the evening shift for the Indian customer support team of a company they later named as Bluevine, a United States-based fintech firm that offers digital banking, business accounts and credit lines for small businesses. According to the post, the team was told that morning to work from home for unspecified operational reasons. Their shift began at 6:30 p.m. local time.

At around 9:30 p.m., senior management sent a calendar invite for a video call. Within about five minutes, the entire 80-person team was told that the Indian customer support operation was being shut down. The employee wrote that no prior warning, performance review or transition period preceded the decision, and that the reason given was that the company was not getting the expected output from the India-based team, a characterisation the poster strongly disputed given years of overtime and high-pressure shifts.

It is important to be precise about what is and is not confirmed here. The story spread through a single anonymous social media account, and Bluevine has not issued a public statement confirming the details, the headcount, or the reasoning. Multiple Indian news outlets, including Gulf News, NDTV-sourced reports and HRKatha, covered the claims but flagged them as employee accounts rather than verified company disclosures. We are treating this the same way: as a widely reported and highly plausible account of a real event, not as an independently proven fact.

What makes the story resonate regardless of the exact details is the pattern it captures. It shows how offshore support and back-office teams, precisely the kind of repetitive, script-driven work that generative AI chatbots and voice agents are now built to handle, can be treated as a cost line that gets deleted in a single meeting rather than a group of careers built over years. Whether or not AI directly replaced these 80 jobs, the manner of the layoff, sudden, unexplained, without a face-to-face conversation, is becoming a recognisable template across the industry, and that is worth taking seriously on its own.

Oracle's 6 A.M. Email: Thousands Gone Before Breakfast

Oracle's layoffs are on much firmer ground, confirmed by CNBC, Reuters, TIME, Forbes and Oracle's own official filings. On the morning of March 31, 2026, employees across the United States, India, Canada, Mexico and Uruguay began receiving termination emails from an address labelled Oracle Leadership before many of them had finished breakfast. The message was short: after careful consideration of Oracle's current business needs, the company had decided to eliminate the recipient's role, and that day was their last working day. System access was cut immediately, with no prior warning from managers.

The exact headcount affected is still not officially confirmed by Oracle, which has repeatedly declined to comment. Reports put the number anywhere from thousands up to a widely cited estimate of 20,000 to 30,000 employees, drawn from an analyst note that had earlier suggested Oracle could save at least eight billion dollars in cash flow by cutting a similar number of roles. Oracle employed about 162,000 people as of its most recent annual filing in May 2025, and a WARN Act filing confirmed that separations tied to this round would continue through June 1, 2026. Some employees, including software engineers and full teams from Oracle's NetSuite division, told TIME they felt they had effectively been asked to train the AI systems that would later replace them, and several described losing hundreds of thousands of dollars in unvested stock overnight.

The financial backdrop explains the pressure Oracle is under. The company has committed to spending roughly fifty billion dollars in capital expenditure in its 2026 fiscal year to build AI data centre capacity, partly through a five-hundred-billion-dollar joint infrastructure project with SoftBank and OpenAI. To fund this, Oracle has leaned heavily on debt markets, raising fifty billion dollars in debt and equity in January 2026 alone, even as its stock price fell around twenty-five percent over the year on concerns about that debt load. Cutting payroll while ramping up capital spending on chips and data centres is, in plain terms, a trade: fewer people, more machines, funded by borrowed money that investors are watching nervously.

Notably, Wall Street's reaction to the layoffs themselves was not fear but relief: Oracle's shares reportedly rose four to six percent on the day the cuts were reported, because investors read the move as a sign of cost discipline. That reaction captures something important about this whole story. What frightens an employee, a sudden email ending a career, can read as good news to a shareholder. Both reactions are rational from where each person is standing, and both are part of the same event.

TCS's 12,200 Job Cuts: India's Largest Tech Layoff Ever

Tata Consultancy Services, India's largest private-sector employer with roughly 613,000 people on its books, announced on July 27, 2025 that it would cut about 12,200 employees, close to two percent of its global workforce, by the end of its 2026 financial year in March. It is the largest downsizing in the company's fifty-year history, and unusually for a move of this scale, it was first revealed in an interview with CEO K. Krithivasan rather than through a formal press release, investor filing or regulatory disclosure to the Bombay or National Stock Exchanges.

TCS's official explanation centres on skills, not AI. The company said the cuts mainly affect middle and senior management whose skills did not match its shift from traditional waterfall software delivery to agile, AI-integrated, product-based models. TCS pointed out that it had already trained more than 550,000 employees in foundational digital skills and 100,000 in advanced skills, but said a segment of the workforce, mostly long-serving employees in legacy roles, could not be repositioned despite those efforts. The company also tightened its bench policy in June 2025, requiring employees between projects to meet minimum utilisation standards, a change that added pressure on exactly the group later affected by the cuts.

Independent observers read the situation differently. Industry analysts and labour reporters covering the story argued that framing the cuts as a skills mismatch understates the role of AI and automation in reducing the actual demand for many of the roles TCS is eliminating, including testing, basic coding and support functions that generative tools now handle at a fraction of the cost. India's Ministry of Electronics and Information Technology was reportedly concerned enough to say it was monitoring the situation and speaking with the company to understand the underlying causes. The layoffs also arrived alongside slower revenue growth at TCS and across India's 283-billion-dollar IT export sector, which has been squeezed by cautious spending from clients in the United States and Europe.

Whichever explanation is closer to the truth, official caution about client spending or unofficial acceleration by AI, the human effect is the same. A single announcement removed more jobs than most Indian cities' entire formal tech sectors, and it did so in an industry that has, for two decades, been one of the most reliable ladders into the Indian middle class.

Meta: Cutting "Low Performers" and Its Own AI Division

Meta's job cuts actually came in two distinct waves that are often blurred together in retellings, so it is worth separating them clearly. The first wave, announced in January 2025 and carried out starting February 10, cut about 3,600 people, five percent of Meta's roughly 72,400-person workforce at the time. CEO Mark Zuckerberg framed this explicitly as a performance-based purge of "low performers," part of what he called an intense year focused on AI, smart glasses and efficiency, not a headcount reduction tied to any specific technology replacing specific jobs. Some employees pushed back publicly, saying they had been cut despite years of strong reviews.

The second wave, in October 2025, was smaller but more directly tied to AI: about 600 roles, roughly a fifth of the 3,000-person Superintelligence Labs division that Meta had built earlier that year around its 14.3-billion-dollar investment in Scale AI and newly hired Chief AI Officer Alexandr Wang. Wang's internal memo said the goal was to strip out what he called organisational bloat inside Meta's own AI research and product teams, mostly within the legacy Fundamental AI Research group, so that decisions could be made faster with fewer people in the loop. Notably, the elite TBD Lab building Meta's most advanced models was left untouched, and Meta invited many of those affected to apply for other roles internally.

The irony is hard to miss: Meta laid off people who worked on AI, from a company that in the same year offered signing packages reportedly worth up to 300 million dollars to lure a small number of top AI researchers away from rivals. Meta's total expenses for 2025 were expected to land between 114 and 118 billion dollars, driven mainly by data centres, chips and computing power rather than by broad-based headcount growth. In other words, Meta is not spending less on AI. It is spending far more on AI infrastructure and elite talent, while spending less on the layers of staff, including some of its own AI workers, that it now considers unnecessary in between.

Microsoft: Five Rounds of Layoffs in a Single Year

No large tech company cut more jobs across more separate announcements in 2025 than Microsoft. The year opened with a small performance-based round in January affecting less than one percent of staff. In May, Microsoft cut about 6,000 people, roughly three percent of its roughly 228,000-person workforce, concentrated in Azure cloud operations, sales and engineering, with software engineering the single largest category hit in its home state of Washington. A further round of at least 300 cuts followed in June. Then, on July 2, 2025, the company cut about 9,000 more people, just under four percent of its global workforce, again citing a need to remove layers of management between individual contributors and senior executives.

By the end of 2025, independent trackers put Microsoft's total job cuts for the year at more than 15,000, close to seven percent of its global headcount, even as the company reported an 18 percent year-over-year jump in quarterly net income to 25.8 billion dollars in the same period it announced the July cuts. The cuts continued into 2026: by July of that year, Microsoft announced a further round primarily targeting its gaming division, which is expected to eliminate 2,850 gaming jobs over the following year as several of its studio brands are spun off or sold, on top of roughly 3,000 employees who had already taken a voluntary retirement package earlier that year.

Microsoft's official language across all of these rounds avoided directly blaming AI for specific job losses, describing the moves instead as "organizational and workforce changes" needed to "position the company and teams for success in a dynamic marketplace." At the same time, CEO Satya Nadella has been unusually candid in public about how much of Microsoft's own code is now written by AI rather than by people, a figure explored in more detail later in this article. Reading the layoffs and that admission together is where much of the public anxiety about "AI taking the jobs" comes from, even though Microsoft itself has never formally attributed a specific layoff number to AI replacing specific roles.

The Bigger Picture: Over a Million US Job Cuts in 2025

Oracle, TCS, Meta and Microsoft are the headline names, but they sit inside a much larger wave. According to Challenger, Gray and Christmas, a global outplacement and workplace data firm that has tracked corporate layoff announcements since the 1980s, US employers announced roughly 1.1 million job cuts across 2025, the highest annual total since the pandemic year of 2020, when more than two million jobs were lost. October 2025 alone saw 153,074 announced cuts, the worst single month since 2003, driven heavily by technology and warehousing.

Two forces dominate this total, and AI is only one of them. Challenger's report found that the single biggest driver in 2025 was what it called the "DOGE Impact," reductions to the US federal workforce and its contractors following the Department of Government Efficiency's cost-cutting drive, accounting for close to 294,000 layoffs on its own. Elsewhere in the world, Accenture cut more than 11,000 jobs in Europe as it retrains staff toward AI agents, Panasonic cut over 10,000 as part of a broader restructuring, and Intel cut between 21,000 and 25,000 amid a 15 to 24 percent global workforce reduction tied to its own turnaround plan. Layoffs.fyi, a separate crowd-sourced tracker, counted more than 180,000 tech job cuts globally through October 2025 across 169 or more companies.

How Much of This Is Really AI? What Challenger Gray Found

This is the number that most directly answers the headline question. Challenger, Gray and Christmas has tracked AI as an explicit, stated reason for layoffs since 2023, and its data lets us see the trend rather than guess at it. In 2025, AI was cited as the reason for about 54,700 to 55,000 planned US job cuts, out of roughly 1.1 million total. That is a real and rapidly growing number, but it represented well under five percent of that year's total layoffs. Since 2023, when companies first started openly blaming AI in layoff notices, the cumulative total attributed to AI passed 71,000 by the end of 2025.

The trend accelerated sharply into 2026. By May 2026, AI accounted for 40 percent of that month's 97,006 announced job cuts, the highest monthly share Challenger had ever recorded, and the year-to-date AI-attributed total for 2026 had already reached 87,714, well ahead of the full-year 2025 figure. Andy Challenger, the firm's chief revenue officer, summarised the finding carefully: AI is not yet the "jobpocalypse" some predicted, but it is now cited more often than any other single reason when companies explain their cuts, ahead of cost-cutting, market conditions, and store or plant closures.

Read this data literally and a clear picture emerges. AI is not yet the dominant cause of job losses in raw numbers. It is, however, the fastest-growing cause, it is spreading from technology into other sectors, and it has overtaken every other single explanation companies give. If the growth rate from 2025 into 2026 continues, AI could become the leading driver of total layoffs, not just the leading cited reason, within a small number of years.

"AI Washing": Are Companies Using AI as an Excuse?

Not everyone accepts these figures at face value, including people who work at the centre of the AI industry itself. OpenAI CEO Sam Altman has publicly said that some companies are "AI washing" their layoffs, using the technology as a convenient explanation for job cuts that are really driven by other business pressures, such as an economic slowdown, past over-hiring during the pandemic boom, or simply a desire to look efficient and innovative to investors. On this view, blaming AI can make a company look forward-looking rather than troubled, even when the underlying cause is more mundane.

Apollo Global Management's chief economist, Torsten Slok, went further, writing that he sees "zero evidence of job losses because of AI" in broader labour market data such as the ADP National Employment Report, which tracks private payrolls across the US economy. His argument is that if AI were truly displacing workers at scale, it should show up clearly in aggregate unemployment and hiring data, not just in the specific reasons individual companies choose to state in press releases.

Both critiques have merit, and they are not actually contradictory. It can be true that AI is a real, growing, and in some cases exaggerated factor at the same time. A company under pressure from investors to control costs has every incentive to attach the word AI to a layoff notice, because it signals strategic vision rather than distress. It can also be true that in specific functions, customer support scripts, routine coding, basic data entry, actual AI tools are now good enough to reduce headcount needs regardless of what a press release says. The honest position is that AI is a real accelerant layered on top of ordinary economic cycles, corporate cost discipline and post-pandemic correction, not the sole cause of the wave, and not a hoax either.

How Much Code Is AI Actually Writing?

One figure gets exaggerated more than any other in retellings of this story: the idea that most of a major tech company's software is now written entirely by AI. The real, well-documented figure is more modest, though still striking. In April 2025, speaking with Meta CEO Mark Zuckerberg at Meta's LlamaCon developer conference, Microsoft CEO Satya Nadella said that inside Microsoft's own code repositories, "maybe 20, 30 percent of the code... is probably all written by software," meaning AI, with results varying by programming language. He said AI performed noticeably better generating Python code than lower-level languages like C++, and that some individual projects were seeing higher shares.

Around the same period, Google CEO Sundar Pichai said on an earnings call that AI was generating more than 30 percent of new code at Google, up from about 25 percent reported roughly six months earlier, suggesting the share is climbing quickly. Zuckerberg, when asked the same question about Meta, admitted he did not know the current figure, but stated a company goal that within about a year, AI might handle roughly half of all development work at Meta, a forward-looking target rather than a reported figure. Microsoft's own chief technology officer, Kevin Scott, has separately said he expects as much as 95 percent of all code to be AI-generated by 2030.

These numbers should be read carefully rather than taken as proof that AI has replaced software engineers wholesale. Companies define "written by AI" differently, and code assisted or autocompleted by a tool is often counted the same as code generated end-to-end without human review. What the figures do show clearly is a real and fast-moving shift inside the daily work of software engineering, one that changes what the job looks like even in companies that are still actively hiring engineers.

Which Kinds of Jobs Are Being Hit First?

Looking across all five case studies together, a pattern emerges in which kinds of work are affected first, and it is not random. Roles that are hit earliest tend to share three features: they involve repeatable, scriptable tasks; their output can be checked relatively easily against a known standard; and they sit in a cost centre rather than a revenue-generating relationship with a client. Customer support, like the Bluevine team, fits this profile closely. So do many mid-level software testing, basic coding and legacy-system maintenance roles at TCS, and many of the management layers Microsoft has targeted in successive rounds.

Roles that have so far been more insulated share the opposite features: judgement calls with high stakes, relationships that clients specifically value, and work where mistakes are expensive or dangerous. That is one reason the flashiest AI-and-jobs headlines about doctors, lawyers and pilots remain mostly speculative rather than documented, even though search interest in those exact questions is high. Regulatory, safety and trust requirements in those fields slow down AI adoption regardless of how capable the underlying models become.

What Research Says: The MIT Estimate

Beyond company announcements, academic research offers a longer-range estimate of AI's potential reach into the labour market. A widely cited study associated with the Massachusetts Institute of Technology estimated that AI could feasibly automate tasks currently performed by around 11.7 percent of the US workforce, representing roughly 1.2 trillion dollars in wages concentrated in sectors including finance, healthcare and professional services. That figure describes technical feasibility, what AI could plausibly do given current capability, not a prediction of how many people will actually be laid off, since adoption speed, cost, regulation and human preference all slow the gap between what is technically possible and what companies actually do.

Surveys of workers themselves show the anxiety is now mainstream rather than confined to tech employees. Reporting tied to Challenger's data cited a meaningful share of US workers who believe their current job is likely to be eliminated within the next five years because of new technology, automation, robots or AI, a sentiment that tracks closely with the search trends shown in the screenshots that prompted this article.

Why Companies Say They Are Doing This

Set aside skepticism for a moment and take the companies' own stated logic seriously, because it is internally consistent even if incomplete. Oracle is spending fifty billion dollars a year building AI data centres and needs cash and organisational focus to do it; cutting payroll frees both. TCS says the market has shifted to agile, AI-assisted delivery, and staff trained for an earlier era of software development genuinely cannot always be redeployed despite real investment in retraining. Meta says its AI division itself became bureaucratic after a hiring spree, and wants fewer layers so that its best researchers can move faster, a claim consistent with its decision to protect its most elite team from the cuts. Microsoft says it is removing management layers to increase agility, a strategy that predates the current AI wave by years but has accelerated alongside it.

None of these explanations amounts to a company saying "we replaced a person with a chatbot," because in most cases that is not literally what happened. What happened, more precisely, is that AI tools raised the output each remaining employee can produce, shifted the skills companies need, and gave executives a fundable, investor-friendly justification for cuts that reduce costs and shift spending toward computing infrastructure instead of people. Whether you call that AI taking the job or a company choosing to spend its money differently is partly a matter of framing, but the practical effect on the person who loses their job is the same either way.

What It Feels Like From the Inside

The numbers above describe policy and strategy. The employee accounts describe something else entirely. Oracle workers told TIME they felt treated like disposable assets, some losing hundreds of thousands of dollars in unvested stock overnight and facing visa-related deportation risk after decades of service. The Bluevine employee wrote about regretting years of unpaid overtime for a company that ended eighty careers in five minutes. TCS employees described the emotional toll of a 45-year-old worker in Kolkata, quoted anonymously by Reuters, who said it was very difficult for someone his age to find new work after being cut loose from the only kind of job he had known.

A recurring detail across nearly every account in this article is the sense of being informed rather than consulted, learning of a decision through an email, a memo, or a five-minute call rather than a conversation that acknowledged years of contribution. That pattern is not unique to AI-era layoffs, mass layoffs have often been handled coldly in the past, but the speed and scale at which it is now happening, often across multiple countries within the same company on the same day, is new, and it is a large part of why these stories go viral rather than staying inside HR departments.

Is Your Job Safe? A Practical Framework

Rather than trying to answer "will AI take my job" as a single yes or no, it helps to break the question into three separate dimensions, because a job can score differently on each one. The first dimension is task repeatability: how much of your daily work follows a predictable pattern that could be described in a written procedure. The second is judgement and accountability: how often your role requires weighing competing considerations, taking legal or safety responsibility, or making a call that a client specifically wants a human to make. The third is relationship value: whether people choose to work with you specifically, rather than with your employer's brand or system in general.

Jobs that score high on repeatability and low on judgement and relationship value, tier-one customer support, basic data entry, routine document review, simple coding tasks, are the most exposed today, and the Bluevine and TCS stories both fit that description closely. Jobs that score low on repeatability and high on judgement or relationship value, complex negotiation, hands-on medical care, courtroom advocacy, high-stakes engineering sign-off, remain comparatively insulated for now, even though AI tools increasingly assist people doing that work rather than replacing them outright.

Warning Signs Worth Watching in Your Own Job

Every company in this article showed recognisable signals before or during its cuts, and workers who noticed them early had more time to prepare. A tightened bench or utilisation policy, as TCS introduced in June 2025, often precedes a layoff round by weeks or months, because it is a mechanism for identifying who to cut. A hiring freeze combined with heavy public spending on AI infrastructure, as at Oracle, signals that a company is choosing capital equipment over payroll growth. Leadership repeatedly using words like "streamline," "remove layers," or "future-ready" in memos, a pattern visible at both Microsoft and TCS, is corporate language that frequently precedes a formal announcement.

At a more personal level, watch for your manager becoming vague about your team's headcount plans, being asked to document your processes in unusual detail, a sudden request to "work from home" without a clear operational reason, similar to what preceded the Bluevine layoff, or your usual one-on-one meetings being cancelled or delegated to someone more junior. None of these signs alone guarantees a layoff, and reacting to every rumour with panic is its own problem, but noticing a cluster of them together is a reasonable trigger to update your resume and quietly test the job market.

What You Can Do Now

The most effective response to this wave is rarely dramatic and almost never involves trying to out-compete AI at the exact task it is best at. Instead, the workers and researchers who study this transition converge on a smaller set of practical moves. Build fluency with the AI tools relevant to your field rather than avoiding them, since being the person who directs and checks an AI system is a more durable position than being the person doing only the task the AI now automates. Move your daily work up the ladder from repeatable execution toward judgement, coordination and client relationships, even in small ways within your current role.

Diversify your professional network beyond your current employer, since several of the layoffs in this article happened with no internal warning at all, and a strong external network shortens the gap between losing a job and finding the next one. Keep a written, up-to-date record of concrete outcomes you have delivered, not just responsibilities you hold, because in a fast layoff you may have very little notice to reconstruct that case for your next employer. Finally, treat your emergency savings as a direct hedge against the kind of five-minute, no-notice termination described in the Bluevine case, since redundancy and severance timelines vary enormously by country and company, as the next section explains.

Severance, Notice Periods and Your Rights

What a company owes a laid-off employee depends heavily on where that employee is based, and the cases in this article illustrate a wide range. In the United States, the WARN Act requires many employers with 100 or more employees to give 60 days' advance notice of a mass layoff or plant closing; Oracle's WARN filing for its March 2026 round confirmed separations continuing through June 1, 2026, and commentary at the time noted that if the company had skipped the required notice period at qualifying sites, affected employees could be owed 60 days of back pay in addition to severance. Oracle offered US employees four weeks of severance plus one additional week for every year of service in an earlier round.

India has no equivalent of the US WARN Act for private-sector white-collar layoffs of this kind, and severance terms are typically governed by individual employment contracts and the Industrial Disputes Act's narrower protections, which mainly cover "workmen" in factory-type roles rather than IT professionals. TCS said it would offer severance packages, career counselling, outplacement support and extended health coverage to affected employees, and pledged senior HR involvement during the transition, though specific amounts were not published. The Bluevine case, if accurate as reported, appears to have involved no severance at all, which is part of why it generated so much public anger.

The practical lesson is simple: know the labour law that applies to your specific location and employment type before you need it, not after, since the range between "60 days of guaranteed notice" and "terminated mid-shift with no notice at all" is enormous, and it is not always predictable from a company's size or reputation alone.

Skills That Are Becoming More Valuable

As routine execution work gets automated, a smaller set of skills is becoming disproportionately valuable precisely because they sit on the other side of the exposure framework described earlier. The ability to specify a problem clearly enough that an AI tool can act on it, sometimes called prompting or AI orchestration, is now a genuine professional skill rather than a novelty. The ability to verify, judge and take responsibility for AI-generated output, whether that is code, a legal draft, or a customer response, is becoming more valuable precisely because AI systems still make confident mistakes.

Domain expertise that lets you spot when an AI-generated answer is subtly wrong is becoming a moat rather than a redundancy, since the people best placed to catch a plausible-sounding but incorrect output are the ones who already deeply understand the underlying field. And soft skills that AI cannot replicate at all, building trust with a client over years, managing a team through a difficult transition, or negotiating a genuinely novel agreement, remain durable precisely because they were never about repeatable execution in the first place.

The Sceptics: Not Everyone Agrees AI Is the Main Driver

It is worth restating the sceptical case in full, because it is easy to lose it under the weight of dramatic headlines. Sam Altman's "AI washing" comment came from the person running the company most associated with the current AI boom, which makes it a notable admission rather than a dismissal from an outsider. Torsten Slok's "zero evidence" comment about broad labour market data is a reminder that individual company press releases are not the same thing as economy-wide statistics, and that companies have every commercial incentive to attach a fashionable explanation to unpopular decisions.

There is also a simpler, less dramatic explanation sitting underneath much of this wave: many large tech companies, including Meta, Microsoft and Amazon, hired extremely aggressively during 2020 to 2022, when low interest rates and pandemic-era demand made growth cheap to fund. Higher interest rates since then, combined with slower growth in some core businesses, created pressure to correct that over-hiring regardless of AI. Add tariff-related costs, persistent inflation, and in India's case a slowdown in discretionary technology spending by Western clients, and a picture emerges in which AI is a real accelerant and a convenient headline, layered on top of an adjustment that was already coming for other reasons.

None of this means the fear behind the search trends shown at the start of this article is misplaced. It means the honest answer to "did AI take the jobs" is neither a full yes nor a full no, but a specific, checkable "partly, and increasingly," a description that will keep shifting as Challenger's own numbers show AI's share of cited reasons climbing sharply from 2025 into 2026.

So, Is AI Taking the Jobs?

Pull the threads of this article together and a fair verdict looks like this. AI did not, on the evidence available, single-handedly cause the layoffs at Oracle, TCS, Meta or Microsoft, and in the Bluevine case the company has not even confirmed AI played any role at all. Each of these companies faced its own specific pressures: Oracle's AI infrastructure debt burden, TCS's shift away from legacy delivery models and softer client demand, Meta's post-hiring-spree correction and internal reorganisation, Microsoft's long-running push to flatten management layers. In every case, however, AI sat visibly in the background, funding the capital spending that competed with payroll, providing the technical capability that made some roles genuinely less necessary, and offering executives and headline writers alike a simple, powerful word to attach to a complicated decision.

The Challenger, Gray and Christmas data gives the clearest single number to hold onto: AI was explicitly the stated reason for about five percent of 2025's US layoffs, a real but modest share, and that share roughly doubled in relative terms within the first half of 2026, becoming the single most frequently cited reason of all. That trajectory, more than any individual company's press release, is the honest answer to the question in the search bar. AI has not taken most of the jobs yet. It is taking a growing, measurable, increasingly visible share of them, and the companies doing the cutting are becoming more comfortable saying so out loud.

Frequently Asked Questions

Did AI really take these jobs at Oracle, TCS, Meta and Microsoft?

Partly. Each company gave its own official reason, ranging from skills mismatches at TCS to management restructuring at Microsoft, and none formally attributed a specific layoff number to AI replacing specific roles. However, all four are simultaneously spending heavily on AI infrastructure or AI-driven delivery models, and independent tracker Challenger, Gray and Christmas found AI was explicitly cited in about 5% of all 2025 US layoffs, rising sharply through 2026.

What actually happened in the Bluevine 80-person layoff?

According to a viral, unverified employee social media post, about 80 Indian customer support staff for the US fintech company Bluevine were reportedly told to work from home, then informed on a five-minute Google Meet call around 9:30 PM that their entire team was being let go. Bluevine has not officially confirmed the details, so this should be treated as a widely reported claim rather than a proven fact.

How many people has Oracle laid off?

Oracle has not officially confirmed an exact number. Reports suggest thousands were let go starting March 31, 2026 via a 6 a.m. email, with some estimates as high as 20,000 to 30,000 out of about 162,000 employees. The cuts came alongside about $50 billion in planned AI infrastructure spending for fiscal 2026.

Why did TCS lay off 12,200 employees?

TCS officially attributed the cuts, announced in July 2025 and completed by March 2026, to a skills mismatch as it shifts from traditional waterfall software delivery to agile, AI-integrated models, mainly affecting middle and senior staff. Independent analysts believe AI and automation played a larger role than the official statement suggests.

Did Meta lay off 8,000 employees because of AI?

Not exactly. Meta's confirmed 2025 layoffs came in two separate rounds: about 3,600 "low performer" cuts in February, described as performance-based rather than AI-driven, and about 600 roles cut from its own AI division in October as part of an internal restructuring, not a combined figure of 8,000.

How many people has Microsoft laid off?

Microsoft cut more than 15,000 employees across multiple rounds in 2025 alone, roughly 7% of its global workforce, including 6,000 in May and 9,000 in July. Further cuts, including about 2,850 gaming roles, were announced in 2026. Microsoft has not officially attributed a specific number of these cuts directly to AI.

How much of Microsoft's and Google's code is written by AI?

Microsoft CEO Satya Nadella said in April 2025 that 20% to 30% of the code in Microsoft's repositories was written by AI. Google CEO Sundar Pichai said AI generated more than 30% of new code at Google around the same time. These are real but partial figures, not evidence that most software is fully AI-written.

What percentage of layoffs are actually caused by AI?

According to Challenger, Gray and Christmas, AI was explicitly cited as the reason for about 54,700 of the roughly 1.1 million US job cuts announced in 2025, under 5% of the total. That share rose sharply in 2026, reaching 40% of layoffs announced in May 2026 alone, making AI the single most cited reason that month.

Are companies using AI as an excuse for layoffs?

Some experts believe so. OpenAI CEO Sam Altman has said companies are "AI washing" layoffs that are really driven by other factors like economic conditions or past over-hiring. Apollo Global Management's chief economist has said he sees no clear evidence of AI-driven job losses in broader labour market data.

Which jobs are most at risk from AI right now?

Roles involving repeatable, scriptable tasks with easily checked output and limited direct client relationships, such as customer support, basic coding, testing and routine data entry, appear most exposed today. Roles requiring high-stakes judgement, safety accountability or personal trust, such as senior medical, legal or engineering sign-off roles, remain comparatively insulated for now.

Is my job safe from AI?

Nobody can say with certainty, but a job is generally more exposed if it is highly repeatable, easily verified against a fixed standard, and part of a cost centre rather than a direct client relationship. Building fluency with AI tools, moving toward judgement-heavy work, and maintaining a strong external network all reduce risk regardless of your specific role.

Conclusion

The viral moments in this story, a five-minute video call, a 6 a.m. email, a fifty-year IT giant's biggest-ever layoff, all point to something real: the balance of power between employers and employees has shifted quickly, and companies now feel comfortable making sudden, large cuts with far less explanation than workers might expect after years of service. Whether AI is the primary cause or a convenient label for cuts that were coming anyway, the practical experience for the people affected is identical, a career interrupted with little warning and, in some cases, little support.

The numbers say AI's direct, admitted role in this wave is still a minority of total layoffs, but a rapidly growing one, already the single most cited reason for job cuts in the most recent data available. That trajectory, not any single company's press release, is the most honest answer to the question people keep typing into Google. The most useful response is not panic, but preparation: understand where your own role sits on the exposure framework in this article, build the AI-adjacent skills that make you harder to replace, know your legal rights before you need them, and keep your financial and professional safety net strong enough to survive a layoff that, as the cases here show, can now arrive with almost no notice at all.