What Is the Impact of AI on Employment? A 2026 Reality Check

Almost every two weeks comes another headline telling us that AI will eliminate hundreds of millions of jobs or AI will generate more jobs than those it will destroy. None of these statements could possibly be entirely correct, and the reality up until now has been much more complicated than these two opposing narratives.

Herein we have compiled everything that has been found out by actual research and reports about how AI will impact the employment situation both globally and in India.

AI Job Predictions at a Glance

Various institutions have arrived at vastly varying figures regarding the effects of AI on employment, all based on different assumptions and timeframes. Here is an analysis of some of the major projections.

Source	Prediction	Scope	Timeframe
World Economic Forum	Net +78 million jobs (170M created vs. 92M displaced)	Global	By 2030
McKinsey Global Institute	20–50 million new jobs; 14% of employees may need to change careers	Global	By 2030
Goldman Sachs	Up to 300 million jobs exposed to automation	Global (US/Europe most affected)	Ongoing
International Monetary Fund	~40% of global employment exposed to AI	Global	Current
MeitY (India)	40–45 million workers redeployed; ~20 million new jobs	India	By 2025
NASSCOM	AI could add $450–500 billion to India’s GDP	India	By 2025

But these figures are not actually contradicting each other because what is meant by “exposure to AI” is not the same as “job losses,” and “new jobs created” is not the same as “jobs created for the same people who lost their jobs.” Take note of that as you continue reading this guide.

What Does the Latest Research Actually Say?

In contrast to media reporting, economists have put together a fairly sophisticated narrative using labor market data. The July 2026 policy brief from the Institute for Economic Policy Research at Stanford (SIEPR) provides one of the most thorough syntheses to date, with the key insight being that the effect of AI on employment has been negligible so far.

Unemployment rates among workers in highly AI-exposed occupations have increased since 2022, but no faster than unemployment rates among the least AI-exposed occupations. Hiring posts in highly exposed occupations such as software development continue to increase. When the U.S. Census Bureau conducted a survey among firms, fewer than 5% of them experienced employment changes due to AI – and half of those said that they gained jobs because of AI and the other half that they lost jobs because of AI.

This does not mean that AI layoffs do not occur – they do, and they are becoming more frequently stated as a reason for them. However, labor economists argue that at least some of these layoffs might be caused by reallocation of money flows towards AI research or an overcorrection following the hiring surge during the pandemic era.

Why Young and Entry-Level Workers Are Feeling It Most

The most telling sign of the impact of artificial intelligence on the labor market is not evident in aggregate figures; rather, it is seen in some specific sectors for recently graduated and entry-level workers. While employment levels for entry-level software developers and customer service representatives dropped after 2022 and especially after 2024, employment for more seasoned workers in those same positions remained stable or even increased.

Young workers have been called canaries in the coal mine by researchers; they are the first to face labor market disruptions as a result of artificial intelligence because junior-level positions tend to include the type of research, writing, and analysis that current AI technologies can perform competently. Whether this effect is the result of AI in particular or other variables, including higher interest rates and post-pandemic hiring patterns, remains to be seen.

How AI Is Changing Existing Jobs

Even where AI isn’t eliminating roles outright, it’s reshaping what those roles look like day to day. The most consistent patterns showing up across industries are:

1. Task automation and role redefinition — rule-based tasks that have been traditionally assigned to people get automated first, while employees increasingly work together with artificial intelligence.
2. New skill requirements — now even non-technical occupations require basic familiarity with artificial intelligence techniques, data analysis, and working with prompts.
3. AI as a productivity multiplier — people use artificial intelligence for performing routine tasks so they can concentrate on decision-making tasks.
4. Shifting industry dynamics — artificial intelligence changes manufacturing into “smart factories,” thus altering the nature of the occupations associated with manufacturing and maintenance.
5. New entrepreneurial opportunities — lower barriers to developing AI applications are driving a startup boom in AI consultancy and narrow AI.
6. A shift from job quantity to job quality — automation of mundane tasks helps to make the rest of the tasks more engaging, not just reduces their number.
7. Constant reskilling— the half-life of particular technical skills decreases, so constant training becomes the norm.
8. New ethical and governance demands — with the rise of AI’s usage in decision-making, such as hiring, issues of fairness, transparency, and bias auditing become real requirements for the job.

Jobs at Risk vs. Jobs Being Created

However, there are no pure “safe” or “at risk” occupations – however, some occupations are definitely more vulnerable to automation compared to other occupations, and another group is rapidly developing due to the development of AI.

More Exposed to Automation	Growing Because of AI
Data entry and basic bookkeeping	AI/ML engineers and developers
Routine customer support (chat/phone)	Data analysts and data scientists
Assembly-line and repetitive manufacturing tasks	AI trainers and data annotators
Entry-level coding, QA, and testing	Prompt and AI-workflow engineers
Basic research, transcription, and first-draft writing	AI ethics, policy, and governance specialists
Routine financial analysis and underwriting	Human-AI collaboration managers
Junior paralegal and document-review work	AI-augmented healthcare, skilled-trade, and technical roles

The pattern to note: roles that are created need skills in AI technology or judgment that cannot be done by AI itself, whereas roles at risk are mechanical in nature and can be defined beforehand.

New Job Roles Emerging From AI

Beyond the well-known “AI engineer” and “data scientist” titles, a wider set of AI-native roles is emerging as adoption spreads:

1. AI Trainers and Teachers – People who are responsible for training AI technology and developing new applications based on it.
2. Data Analysts and Scientists – Specialists who analyze the vast amounts of data generated by AI and make recommendations for businesses.
3. Human-Machine Teaming Managers – People responsible for collaboration between AI systems and human employees specifically.
4. AI Ethics and Policy Specialists – Professionals who make sure that the development of AI technologies is done in an ethical manner.
5. Workflow Engineers – Specialists who define how organizations can make use of AI technology.
6. AI Auditors and Red-Teamers – This relatively new type of job includes people responsible for testing AI technology for any biases and vulnerabilities.

AI’s Impact by Sector

AI adoption and its employment effects vary significantly by industry. Here’s a snapshot of where things stand across the sectors most commonly discussed.

Sector	Main AI Use	Employment Effect So Far
IT & Software	Coding assistants, automated testing	Mixed — productivity gains for experienced devs, softer hiring for entry-level roles
Customer Service	Chatbots, virtual assistants	Routine-query roles shrinking; complex-case handling still human-led
Healthcare	Diagnostic tools, imaging analysis, telemedicine, AI scribes	Augmenting rather than replacing — clinician oversight still required
Finance	Fraud detection, credit underwriting, risk assessment	Growing demand for AI-literate finance and risk specialists
Manufacturing	Robotics, smart-factory systems, predictive maintenance	Slower disruption than expected — human labor still more flexible for many tasks
Retail & E-commerce	Personalization, demand forecasting, inventory automation	Efficiency gains; some reduction in routine merchandising roles
Agriculture	Precision farming, yield prediction	Least automated sector so far; large potential but low current adoption

Does AI Actually Make Workers More Productive?

The other side of the employment picture is productivity, and it is more consistently good than one might expect from the job loss perspective – although its benefits also differ according to skills level and tasks. One of the best-known experiments on developer productivity found that using the GitHub Copilot assistant allowed completing coding tasks 56% faster, with novice programmers getting the most benefit.

In another experiment involving a call center with thousands of operators, the use of an AI assistant boosted productivity by 15%, but almost entirely because of inexperienced workers – skilled agents did not see any improvements. Similarly, writing tools demonstrated that AI assistance made employees work faster and better mostly because of their lack of skill.

Scientists speak of “jagged” productivity of AI – very helpful in some tasks and even unhelpful in others, depending on their compatibility with the strengths of artificial intelligence. This means that the ability to benefit from the use of AI will require learning to recognize its limitations and knowing when to use it and when to refrain from doing so.

The Global Picture: US, UK, and Europe

AI’s exposure to disruption in major economies of the West is very hard to quantify and varies a lot depending on methodology, but the trend is always clear: office work involving complex cognitive functions is most vulnerable. Goldman Sachs predicts up to 300 million people worldwide might get disrupted by the technology, with almost two thirds of exposed jobs located in the United States and Europe.

For example, a recent poll in Britain included 22,000 job types, revealing AI’s impact on 8 million jobs, with 11% of tasks currently vulnerable to automation. Based on its research, McKinsey forecasts AI and automation will force 14% of global workforce to change their occupations by 2030.

However, despite such big numbers, data from employers is much more restrained. A mere fraction of firms asked by the Census Bureau to reveal the effects of AI on their employees reports any changes, and firms that have already implemented AI have increased, not decreased, employment within two years after implementation.

AI and Employment in India

The Indian experience with employment in AI has its own contours based on the sheer scale of its IT industry and its huge informal workforce. The IT industry employs more than 5.4 million people. Historically, it has been one of the biggest employers for engineering graduates. However, the automation of even the simpler programming and testing activities is leading to a decrease in the number of jobs in this field – leading to predictions of a potential slowdown in white collar jobs by 2027.

The manufacturing industry too has its own challenges – according to a McKinsey report, automation can lead to displacement of 60 million manufacturing jobs in India by 2030 – mainly in labor-intensive industries such as textiles and electronics. Around 300 million workers in blue-collar occupations in manufacturing and related industries could be impacted in some way or another.

A second larger structural challenge is posed by the Indian informal workforce, which constitutes 90% of the total workforce, who do not have any formal agreements, social security, or avenues for re-skilling and hence are more susceptible to disruptions with fewer safety nets. On the positive side, according to the WEF forecast, India will experience a net addition of 12 million jobs by 2025, and according to the government estimation, the digital revolution through AI can redeploy 40-45 million people to create an estimated 20 million jobs.

India’s Policy Response

India’s government has rolled out several initiatives specifically aimed at managing this transition:

1. Skill India Mission/PMKVY – leading national initiative for skills development that provides training in artificial intelligence, machine learning, robotics, and data analysis.
2. Digital India Mission – an initiative that is aimed at improving overall digital literacy and technology usage.
3. Atal Innovation Mission – provides funding for establishment of Atal Tinkering Labs in schools and Atal Incubation Centres to help AI-based startups.
4. NITI Aayog’s National Strategy for Artificial Intelligence – focuses on use of AI for good in healthcare, agriculture, education, smart cities, and transportation.
5. National Education Policy 2020 – encourages more integration of AI skills in the educational system.
6. FutureSkills PRIME – platform for lifelong learning that includes emerging technologies such as AI.
7. YUVAi – teaches students from class 8 to 12 about AI.

Skills That Will Matter Most in an AI Economy

Generic advice to “reskill” isn’t very actionable. Based on where labor-market demand is actually shifting, these are the specific capabilities worth prioritizing:

1. AI/data literacy – an understanding of what an AI tool can and cannot do beyond the technical profession.
2. Prompt and Tool fluency – actual skills to generate meaningful results from an AI tool within your professional context.
3. Judgment/AI-output evaluation – recognizing whether to trust an AI result or not and how to fix an error if there is one.
4. Deep domain knowledge along with AI tools – having deep domain knowledge along with AI literacy as opposed to just one or the other.
5. Critical thinking and problem solving – skills that an AI tool cannot mimic as accurately as humans can.
6. Creativity – particularly creativity that results in unique/original ideas as opposed to AI results.
7. Emotional intelligence/communication – relationship-based and persuasive work is difficult to automate.
8. Adaptability and continuous learning – skills to always update yourself based on evolving tools/workflows.

How to Future-Proof Your Career in the AI Era

Beyond building the right skills, a few concrete habits make the biggest difference in staying resilient as AI reshapes your industry:

1. Know the AI technologies relevant to your particular field – not AI as such, but those AI technologies relevant to your actual profession.
2. Develop your proficiency in data and AI – not by waiting for it to become mandatory, but by taking classes, getting certified, or working on projects.
3. Develop human traits – judgement, creativity, relationship building and leadership cannot be automated and are highly valued these days.
4. Keep track of how AI changes your role – based not on your assumption, but on industry trends and your own company’s developments.
5. Develop your portfolio of judgement tasks – those which demand expertise and context beyond mere execution.
6. Stay connected – network, because career opportunities will come through relationships before new job titles reflect the change.
7. Make learning an ongoing effort, not a class – because skills that matter now will continue changing in the future.

Frequently Asked Questions

Conclusion

When it comes to answering the question about the effects of artificial intelligence on jobs, the truth is that it’s actually less dramatic than the most frightening reports, but also more real than those which paint the most rosy picture. There have not been mass layoffs caused by artificial intelligence yet, but workers in low-end jobs in occupations threatened by it have already felt its influence, which is something to pay attention to, instead of ignoring it.

The way ahead seems clear no matter where you work, whether in Mumbai or Manchester. You need to develop skills related to using AI technologies in your area, learn to make use of the capabilities you have that cannot be replicated by AI, and keep retraining yourself in a constant manner, instead of doing it only once.