The numbers are real. In 2025, nearly 245,000 tech jobs were eliminated globally, with approximately 70% at US-headquartered companies. In the first two months of 2026 alone, the pace has accelerated: 45,000 tech layoffs by March 9, with over 9,200 explicitly attributed to AI and automation. Block cut 4,000 employees. Amazon eliminated 30,000 corporate roles in two separate rounds — about 10% of its corporate workforce. WiseTech Global restructured with 2,000 layoffs, citing AI-driven productivity improvements so dramatic that 'traditional approaches to writing and maintaining code are becoming increasingly obsolete.' A Stanford study confirmed what many had feared: developers aged 22–25 have lost nearly 20% of their jobs since ChatGPT launched in late 2022, with employment for older developers remaining stable or growing.
If you are reading those numbers with a knot forming in your stomach — you are having a rational response. But you are also, very likely, making cognitive errors that will lead you to bad decisions. The loudest story about AI and employment is the catastrophe story: universal replacement, mass unemployment, the end of meaningful human work. It is a story that sells well, generates clicks, drives anxiety, and — according to Forrester Research, CNN Business, and every serious labour economist tracking the data — substantially overstates what is currently happening while underpreparing people for what is actually coming. The real story is more nuanced, more demanding, and more navigable than the headlines suggest. Let us examine it with the precision it deserves.
What Is Actually Happening: Three Separate Phenomena Being Conflated
The 2026 job market narrative conflates three distinct phenomena that have very different causes and very different implications. Separating them is the first step toward clear thinking.
Phenomenon 1: The Pandemic Hiring Correction
Between 2020 and 2022, tech companies overhired massively in response to pandemic-driven digital demand. When that demand normalised and interest rates rose dramatically, companies found themselves with bloated headcounts built for an economy that no longer existed. The layoffs from 2022 through early 2025 were substantially — perhaps primarily — the correction of this overhiring, not AI displacement. According to Indeed FRED data, software engineering job postings peaked in mid-2022 and have since declined to below pre-pandemic levels. This is partly normalisation, partly rate-driven caution, and only partly AI.
Phenomenon 2: The AI Cover Story
Forrester Research's 2026 Future of Work report contains a striking finding: 55% of employers report regretting AI-attributed layoffs, and according to Forrester, half of AI-attributed layoffs will be quietly rehired — but offshore, at significantly lower salaries. The pattern, Forrester argues, is not primarily AI replacing humans — it is AI providing executives with a narrative to justify cost-cutting decisions that are primarily driven by margin pressure and investor expectations. When a company lays off workers 'for AI' and then quietly fills the gaps with offshore contractors at lower cost, AI is not the cause of the displacement. Cost arbitrage — enabled by the normalisation of remote work during the pandemic — is. AI is the story told to investors.
Phenomenon 3: Real AI Displacement of Specific Roles
There is genuine AI-driven displacement, and it is not evenly distributed. The Burning Glass Institute finds that AI tools like GPT can instantly handle foundational tasks that once taught early-career workers the fundamentals of their profession: drafting, basic analysis, initial research, documentation. Employers conclude they do not need junior staff if AI can do the work. They flatten organisational structures and hire mid-career professionals who require less training. The Stanford study's finding — that young software developers (22–25) have lost 20% of roles while senior developers remain stable — is the clearest evidence of this structural shift. The displaced positions are not being replaced by AI performing autonomous software engineering. They are being replaced by companies requiring fewer humans to manage AI-assisted development workflows.
The Entry-Level Crisis: Why This Matters Deeply for Indian Students
For Indian engineering students graduating in 2026, the entry-level displacement is the most practically significant dimension of the job market story. The traditional pathway — graduate, take a junior role, learn on the job over three years, advance — is under genuine pressure. Companies that previously hired 10 junior developers to do the work now hire 3 mid-level developers who use AI tools to maintain the same throughput. The junior roles that served as apprenticeships are vanishing faster than the mid-level and senior roles.
GitHub CEO Thomas Dohmke stated this plainly on X: developers must 'either embrace AI or get out of this career.' Anthropic CEO Dario Amodei, in a widely discussed message specifically directed at Indian engineers, advised focusing on understanding AI systems rather than treating coding as a terminal skill — the ability to direct AI effectively matters more than the ability to write every line manually. These are not abstract philosophical positions. They are descriptions of what hiring managers are actually rewarding.
The data confirms this. According to InformationWeek's 2026 layoff tracker, companies posting the most aggressive AI-driven cuts — Block, WiseTech, eBay, Oracle — are simultaneously posting job openings for AI engineers, ML infrastructure specialists, and LLM product managers. The job market is not contracting uniformly. It is bifurcating: roles that work with AI are growing; roles that AI can replicate are shrinking. The students who understand this bifurcation and position themselves on the right side of it will enter a job market of genuine opportunity. Those who do not are entering a market that will feel increasingly hostile.
What the Doom Forecasts Get Wrong
The most extreme job market forecasts — Roman Yampolskiy's claim that 99% of workers will be jobless by 2030, the Citrini Research viral essay projecting US unemployment above 10% by 2028 — are not forecasts built on labour market data. They are thought experiments about a hypothetical AI capability trajectory extrapolated to social outcomes without accounting for friction: institutional inertia, regulatory lag, deployment complexity, consumer trust, and the simple fact that only 11% of organisations have agentic AI systems in production right now, according to Deloitte.
CNN Business's analysis, published March 2026, put it precisely: 'The labor market has been cooling, though layoffs remain at levels economists consider historically manageable. The unemployment rate in January was at 4.3% — about half a percentage point higher than late 2023, when the generative AI boom began. Automation has been buffeting the labor market for decades and it hasn't led to universal structural collapse. It is also worth noting that the most dire forecasts often come from the very executives who stand to profit the most from selling the AI products they claim are transformative.' The doom loop narrative — AI replaces workers, companies shed jobs, profits rise, more AI investment, more layoffs — is coherent in theory and absent in current data.
The New Skills That Are Actually Commanding Premiums
What does the job market actually reward in March 2026? The evidence from hiring data, salary surveys, and Forrester's AIQ research is consistent. Roles at the intersection of domain expertise and AI fluency are growing fastest and commanding the highest salaries. An accountant who understands AI-driven audit tools commands significantly more than one who does not. A bioinformatician who can fine-tune LLMs on genomic data is extraordinarily sought-after. A software engineer who can architect multi-agent systems and evaluate LLM outputs commands 40–60% premium over one who cannot.
The Forrester research on AIQ — artificial intelligence quotient, their measure of AI readiness — found that only 16% of individual workers had high AIQ in 2025, projected to rise to just 25% in 2026. The majority of organisations are not investing in AI training: only 23% of AI decision-makers report that their organisations offered prompt engineering training in 2025. Employees are teaching themselves through solo experimentation. This is a window of opportunity for the students and professionals who actively build AI fluency now, before it becomes a standard expectation. The scarcity of high-AIQ workers is precisely what commands the premium.
India-Specific Dynamics: Why the Story Is Different Here
The Indian job market has unique characteristics that make the AI disruption story simultaneously more complex and, in some respects, more manageable than the US picture. The IT services sector — which employs millions of Indian engineers in relatively routine software maintenance, QA, and support roles — is the most exposed to AI substitution. Companies like WiseTech that explicitly cited 'traditional approaches to writing and maintaining code are becoming obsolete' are describing exactly the kind of work that significant portions of India's IT services workforce perform.
At the same time, India's domestic AI economy is growing rapidly. NASSCOM reports AI-related job postings in South Asia grew from 2.9% to 6.5% of total vacancies between 2023 and 2025, with demand growing 75% faster than non-AI roles. India's GCC ecosystem — 1,700+ Global Capability Centres employing over 1.9 million people — is actively upgrading to AI-first workflows, creating demand for exactly the Indian engineers who understand AI systems. The TCS-Infosys-Wipro tier is contracting at the routine end while the GCC and product company tier is growing at the AI-fluent end. The movement between these tiers is real and available to students who prepare for it.
The Honest Framework for Career Decisions in 2026
Here is what the evidence, stripped of both panic and false comfort, actually supports. First: AI is genuinely displacing routine, well-defined, codified tasks across all knowledge work domains. This is not hype and not reversible. Second: the displacement is concentrated at the entry level and in routine roles, not at the senior and specialised end. Third: new roles at the AI-human intersection are growing faster than routine roles are contracting, but they require different skills and the transition is not automatic. Fourth: the organisations claiming AI is replacing entire departments are often overstating for investor audiences and will quietly rehire half those roles within 24 months. Fifth: the students and professionals who build genuine AI fluency now — not surface-level familiarity but working depth — will operate in a fundamentally better market than those who do not.
The honest career advice for 2026 is not 'panic and change everything' nor 'AI is hype and nothing will change.' It is: understand which specific tasks in your intended field AI does well, which it does poorly, and where the combination of human judgment and AI capability creates irreplaceable value. Position yourself at that intersection. Build AI fluency as a practised skill, not a listed credential. And recognise that the window for doing this before it becomes universal expectation — and therefore stops commanding a premium — is probably two to three years.
Pro Tip: The most useful thing you can do this week: identify the three to five most routine, AI-replicable tasks in your intended career. Then identify the three to five tasks in that career that require contextual judgment, ethical reasoning, relationship management, or physical presence. The second list is your career foundation for the next decade. The first list is what you should be learning to manage with AI tools, not protecting as a job description.