Why read this: Big tech firms (Atlassian, Block, Amazon, Meta) have all linked recent staff cuts to AI, and that explanation is travelling fast through news feeds your students already read. This article is worth assigning because it teaches them to interrogate that narrative: it concedes a real but narrow effect from AI in specific occupations (computer programmers, customer service, data entry, call centres), then sets that against three competing drivers (corporate restructuring, over-hiring, investor signalling) and a striking counter-example in Meta's US$600 billion AI commitment. For Mandarin L1 readers heading into university or first jobs in AI-exposed sectors, the piece models how to weigh evidence against rhetoric on a story that directly affects their futures.
What to notice: Watch how the writer builds the argument by concession rather than attack. Paragraph 1 lays out the corporate narrative; the single sentence 'The evidence, however, tells a more nuanced story' pivots the entire piece. Notice the density of nominalisation in paragraphs 5 and 6 ('a workforce reduction framing built around AI adoption', 'subsidising the AI bet') and the way each abstract noun phrase carries an implicit verb-clause underneath it. Track the hedges (tentative, telling, plausible, would have been made) and the quiet authority claims through named research bodies (Anthropic, Goldman Sachs, PwC). The closing 'conflate' sentence asks readers to hold two competing causal stories in working memory at once.
Skills practised: Reading: tracking an implicit thesis through concede-then-pivot structure, decoding nominalised noun phrases back into verb clauses, interpreting hedged modality as analytical caution rather than weakness, and synthesising statistics across three cited reports. Writing: practising hedged claims with may, suggests, tends to and likely; deploying discourse markers (however, moreover, while, at the same time) to signal argumentative pivots; and writing a one-sentence summary frame that captures both the corporate narrative and the writer's counter-claim.
Tech companies are blaming massive layoffs on AI. What is really going on?
Atlassian, Block, Amazon and Meta all point at AI when they cut jobs. The research suggests several other forces are doing more of the work.
Tap any green word in the article to see its meaning.
When tech corporations announce significant staff cuts: reductions in the number of employees a company keeps on its payroll — 裁员, the explanation is usually short. Atlassian, Block and Amazon have all said they would lay off: to dismiss employees, usually because the company no longer needs them — 解雇 thousands of employees, and have attributedattribute: To ascribe (something) to a given cause, reason etc — 归因于；归咎于 the cuts to efficiency gains: improvements in productivity that let the same work be done with fewer resources — 效率提升 delivered by artificial intelligence (AI). The narrative is consistent: AI is making human labour replaceable, and responsible managers are simply adjusting. The evidence, however, tells a more nuancednuanced: Having nuances; possessed of multiple layers of detail, pattern, or meaning — 细致入微的 story.
Genuine disruptiondisruption: A disturbance or problem that interrupts an activity, event, or process. — 干扰；中断；混乱 is visible in specific corners of the labour market: the supply of and demand for paid work in an economy — 劳动力市场, though its scale is commonly overstated. Research published this month by Anthropic shows that although many tasks are susceptiblesusceptible: Likely to be affected by something — 易受影响的 to automation, the vast majority are still performed primarily by humans. Some occupations are more exposed to displacementdisplacement: The act of displacing, or the state of being displaced; a putting out of place — 移置；流离失所；使脱离原位的行为或状态 than others (computer programmers top the list, followed by customer service representatives and data entry clerks), yet even within these jobs, AI use remains limited.
The aggregate economic picture reflects this restraintrestraint: the act of holding back from action, or a measure that limits or controls freedom — 克制；约束；限制. A 2025 Goldman Sachs report estimated that, if AI were used everywhere it currently could be, roughly 2.5% of US employment would be at risk of: in danger of (a bad outcome) — 面临……风险 job loss: the situation of becoming unemployed — 失业. That is no trivial number. The same report notes, however, that workers in AI-exposed: (of jobs/sectors) likely to be affected by automation through AI — 易受AI冲击的 occupations are no more likely than anyone else to face reduced hours or lower wages today. The report does flag early signs of strain in narrower pockets, including marketing, graphic design, office administration and call centres: offices where staff handle large volumes of customer phone calls — 客服中心.
Younger workers face sharper pressure. In the first half of 2025, US tech workers in their 20s in AI-exposed roles saw unemployment rise by almost 3%. Anthropic's data shows job-finding rates: the share of unemployed people who get a job within a set period — 再就业率 for those aged 22 to 25 entering such occupations have fallen by around 14% since ChatGPT launched in 2022. These signals are real, tentative but telling, yet they are sector-specific and concentrated, not the sweeping displacement: wide-scale loss of jobs across many industries — 大规模岗位流失 that corporate announcements often imply. The gap between the rhetoricrhetoric: the art of effective or persuasive speaking or writing, often implying a lack of sincerity or meaningful content — 修辞；言辞；花言巧语 and the evidence warrantswarrant: To justify; to give grounds for — 正当 closer examination.
Several other forces are running at the same time as genuine advances in AI: corporate restructuring: major reorganisation of a company, often involving layoffs — 公司重组, over-hiring: the practice of recruiting more staff than a company actually needs — 过度招聘 during the post-pandemic boom: the surge in business activity that followed the COVID-19 pandemic — 疫情后繁荣期 as demand for online services soaredsoar: To rise, especially rapidly or unusually high — 飙升, and pressure from investors to demonstrate improved profit margins: the share of revenue that remains as profit after costs — 利润率. These explanations are not mutually exclusive: such that if one is true the others cannot also be true — 互斥的, but they are rarely acknowledged together in corporate communications. There is also a powerful incentive to be seen embracing AI aggressively. Since ChatGPT launched, AI-related stocks have accounted for about 75% of S&P 500 returns: the gains earned by the 500 largest US-listed companies — 标普500指数回报, and a workforce reduction: a cut in the number of people a company employs — 裁员 framing built around AI adoption: a company starting to use AI tools in its operations — 采用AI sends a different signal to markets than a blunt cost-cutting: aimed at reducing a company's expenses — 削减成本的 announcement does. A firm pursuing AI innovation looks healthier than one sackingsack: To discharge from a job or position; to fire — 解雇 staff because of declining revenues: falling income from sales over time — 营收下滑.
It is worth distinguishing two kinds of cut. In the first, AI genuinely raises productivityproductivity: the rate or efficiency at which goods or services are produced, relative to the resources used — 生产力；生产率 to the point where fewer workers are needed. In the second, the cuts fund the AI itself. Meta illustrates this neatly: the company is reportedly planning to lay off as much as 20% of its workforce while simultaneouslysimultaneously: at the same time — 同时地；同步地 committing US$600 billion to build data centres: large facilities that house computer servers, often used to run AI — 数据中心 and recruit top AI researchers. The workers being let go are not being replaced by AI today; they are subsidisingsubsidise: To assist (someone or something) by granting a subsidy — 补贴 the AI bet: a large risky investment a company makes on AI paying off — AI豪赌 their employer is making on the future.
The big picture: the overall situation, ignoring small details — 总体形势 is more plausible as transformationtransformation: a thorough or dramatic change in form, appearance, or character — 转变；变革；转型 than as elimination. A recent PwC report finds that employment is still growing in most industries exposed to AI, although growth tends to be slower than elsewhere. At the same time, wages in those industries are rising about twice as fast as in the least affectedaffected: Influenced or changed by something — 受影响的 sectors, and workers with AI skills command an average wage premium: the extra pay earned for having a particular skill — 工资溢价 of around 56%. Together the data points to a flatteningflattening: The act, or the result of making something flat of flatter — 平坦化 of the traditional workplace pyramid: the hierarchical structure of employees from junior to senior — 职场金字塔 rather than mass displacement: large-scale loss of jobs across the workforce — 大规模失业: fewer junior employees doing routine analyticalanalytical: Of or pertaining to analysis; resolving into elements or constituent parts — 分析的 and administrativeadministrative: Relating to the management or organisation of a business, institution, or government. — 行政的；管理的 work, while experienced professionals who deploydeploy: To move troops or equipment into position for action; to use or employ resources effectively. — 部署；调配；有效利用 AI tools effectively become more productiveproductive: achieving a significant amount of output or results; yielding favourable outcomes — 高产的；富有成效的.
AI is a consequentialconsequential: Having significant consequences; of importance — 重大的 technology, and its long-term trajectory will reshape work. What is in doubt is whether the dramatic, AI-attributed staff cuts announced by individual companies accurately reflect that path, or whether they conflateconflate: to (by extension) To fail to properly distinguish or keep separate (things); to mistakenly treat (them) as equivalent — 混淆 genuine technological change with decisions that would have been made regardless. Drawing the line is not an academic exercise. It shapes how policymakerspolicymaker: One involved in the formulation of policies, especially politicians, lobbyists, and activists — 决策者, educatorseducator: A person who provides instruction or education, especially as a profession. — 教育工作者；教育家 and the workers themselves understand the disruption they are navigatingnavigate: To find a way through a difficult situation or process — 应对.
When tech corporations announce significant staff cuts: reductions in the number of employees a company keeps on its payroll — 裁员, the explanation is usually short. Atlassian, Block and Amazon have all said they would lay off: to dismiss employees, usually because the company no longer needs them — 解雇 thousands of employees, and have attributedattribute: To ascribe (something) to a given cause, reason etc — 归因于；归咎于 the cuts to efficiency gains: improvements in productivity that let the same work be done with fewer resources — 效率提升 delivered by artificial intelligence (AI). The narrative is consistent: AI is making human labour replaceable, and responsible managers are simply adjusting. The evidence, however, tells a more nuancednuanced: Having nuances; possessed of multiple layers of detail, pattern, or meaning — 细致入微的 story.
Genuine disruptiondisruption: A disturbance or problem that interrupts an activity, event, or process. — 干扰；中断；混乱 is visible in specific corners of the labour market: the supply of and demand for paid work in an economy — 劳动力市场, though its scale is commonly overstated. Research published this month by Anthropic shows that although many tasks are susceptiblesusceptible: Likely to be affected by something — 易受影响的 to automation, the vast majority are still performed primarily by humans. Some occupations are more exposed to displacementdisplacement: The act of displacing, or the state of being displaced; a putting out of place — 移置；流离失所；使脱离原位的行为或状态 than others (computer programmers top the list, followed by customer service representatives and data entry clerks), yet even within these jobs, AI use remains limited.
The aggregate economic picture reflects this restraintrestraint: the act of holding back from action, or a measure that limits or controls freedom — 克制；约束；限制. A 2025 Goldman Sachs report estimated that, if AI were used everywhere it currently could be, roughly 2.5% of US employment would be at risk of: in danger of (a bad outcome) — 面临……风险 job loss: the situation of becoming unemployed — 失业. That is no trivial number. The same report notes, however, that workers in AI-exposed: (of jobs/sectors) likely to be affected by automation through AI — 易受AI冲击的 occupations are no more likely than anyone else to face reduced hours or lower wages today. The report does flag early signs of strain in narrower pockets, including marketing, graphic design, office administration and call centres: offices where staff handle large volumes of customer phone calls — 客服中心.
Younger workers face sharper pressure. In the first half of 2025, US tech workers in their 20s in AI-exposed roles saw unemployment rise by almost 3%. Anthropic's data shows job-finding rates: the share of unemployed people who get a job within a set period — 再就业率 for those aged 22 to 25 entering such occupations have fallen by around 14% since ChatGPT launched in 2022. These signals are real, tentative but telling, yet they are sector-specific and concentrated, not the sweeping displacement: wide-scale loss of jobs across many industries — 大规模岗位流失 that corporate announcements often imply. The gap between the rhetoricrhetoric: the art of effective or persuasive speaking or writing, often implying a lack of sincerity or meaningful content — 修辞；言辞；花言巧语 and the evidence warrantswarrant: To justify; to give grounds for — 正当 closer examination.
Several other forces are running at the same time as genuine advances in AI: corporate restructuring: major reorganisation of a company, often involving layoffs — 公司重组, over-hiring: the practice of recruiting more staff than a company actually needs — 过度招聘 during the post-pandemic boom: the surge in business activity that followed the COVID-19 pandemic — 疫情后繁荣期 as demand for online services soaredsoar: To rise, especially rapidly or unusually high — 飙升, and pressure from investors to demonstrate improved profit margins: the share of revenue that remains as profit after costs — 利润率. These explanations are not mutually exclusive: such that if one is true the others cannot also be true — 互斥的, but they are rarely acknowledged together in corporate communications. There is also a powerful incentive to be seen embracing AI aggressively. Since ChatGPT launched, AI-related stocks have accounted for about 75% of S&P 500 returns: the gains earned by the 500 largest US-listed companies — 标普500指数回报, and a workforce reduction: a cut in the number of people a company employs — 裁员 framing built around AI adoption: a company starting to use AI tools in its operations — 采用AI sends a different signal to markets than a blunt cost-cutting: aimed at reducing a company's expenses — 削减成本的 announcement does. A firm pursuing AI innovation looks healthier than one sackingsack: To discharge from a job or position; to fire — 解雇 staff because of declining revenues: falling income from sales over time — 营收下滑.
It is worth distinguishing two kinds of cut. In the first, AI genuinely raises productivityproductivity: the rate or efficiency at which goods or services are produced, relative to the resources used — 生产力；生产率 to the point where fewer workers are needed. In the second, the cuts fund the AI itself. Meta illustrates this neatly: the company is reportedly planning to lay off as much as 20% of its workforce while simultaneouslysimultaneously: at the same time — 同时地；同步地 committing US$600 billion to build data centres: large facilities that house computer servers, often used to run AI — 数据中心 and recruit top AI researchers. The workers being let go are not being replaced by AI today; they are subsidisingsubsidise: To assist (someone or something) by granting a subsidy — 补贴 the AI bet: a large risky investment a company makes on AI paying off — AI豪赌 their employer is making on the future.
The big picture: the overall situation, ignoring small details — 总体形势 is more plausible as transformationtransformation: a thorough or dramatic change in form, appearance, or character — 转变；变革；转型 than as elimination. A recent PwC report finds that employment is still growing in most industries exposed to AI, although growth tends to be slower than elsewhere. At the same time, wages in those industries are rising about twice as fast as in the least affectedaffected: Influenced or changed by something — 受影响的 sectors, and workers with AI skills command an average wage premium: the extra pay earned for having a particular skill — 工资溢价 of around 56%. Together the data points to a flatteningflattening: The act, or the result of making something flat of flatter — 平坦化 of the traditional workplace pyramid: the hierarchical structure of employees from junior to senior — 职场金字塔 rather than mass displacement: large-scale loss of jobs across the workforce — 大规模失业: fewer junior employees doing routine analyticalanalytical: Of or pertaining to analysis; resolving into elements or constituent parts — 分析的 and administrativeadministrative: Relating to the management or organisation of a business, institution, or government. — 行政的；管理的 work, while experienced professionals who deploydeploy: To move troops or equipment into position for action; to use or employ resources effectively. — 部署；调配；有效利用 AI tools effectively become more productiveproductive: achieving a significant amount of output or results; yielding favourable outcomes — 高产的；富有成效的.
AI is a consequentialconsequential: Having significant consequences; of importance — 重大的 technology, and its long-term trajectory will reshape work. What is in doubt is whether the dramatic, AI-attributed staff cuts announced by individual companies accurately reflect that path, or whether they conflateconflate: to (by extension) To fail to properly distinguish or keep separate (things); to mistakenly treat (them) as equivalent — 混淆 genuine technological change with decisions that would have been made regardless. Drawing the line is not an academic exercise. It shapes how policymakerspolicymaker: One involved in the formulation of policies, especially politicians, lobbyists, and activists — 决策者, educatorseducator: A person who provides instruction or education, especially as a profession. — 教育工作者；教育家 and the workers themselves understand the disruption they are navigatingnavigate: To find a way through a difficult situation or process — 应对.
Questions
Check your understanding
- 01
Which statement best captures the writer's central argument about AI-attributed layoffs?
- 02
What does the Meta example (laying off 20% of workers while committing US$600 billion to AI infrastructure) most directly illustrate?
- 03
Why does the writer return to wages and the 56% wage premium near the end of the article?
- 04
Assess the claim that companies attribute layoffs to AI primarily for investor signalling rather than because AI has displaced the workers in question. Use at least two specific pieces of evidence from the article.
Suggested length: ~100 words
- 05
Argue whether the writer's concluding distinction (between genuine technological change and decisions that would have been made regardless) matters for policymakers and educators. Defend your position with reference to the article.
Suggested length: ~100 words
Questions
Check your understanding
- 01
Which statement best captures the writer's central argument about AI-attributed layoffs?
- 02
What does the Meta example (laying off 20% of workers while committing US$600 billion to AI infrastructure) most directly illustrate?
- 03
Why does the writer return to wages and the 56% wage premium near the end of the article?
- 04
Assess the claim that companies attribute layoffs to AI primarily for investor signalling rather than because AI has displaced the workers in question. Use at least two specific pieces of evidence from the article.
Suggested length: ~100 words
- 05
Argue whether the writer's concluding distinction (between genuine technological change and decisions that would have been made regardless) matters for policymakers and educators. Defend your position with reference to the article.
Suggested length: ~100 words