Artificial intelligence is accelerating the transformation of work and is no longer a futuristic concept: it's the daily reality in companies and public administrations. Between automation, new data flows, and generative tools, AI has permeated key processes and changed the rules of the game for millions of professionals. In this context, Understanding magnitudes, sectors, and policies It's not a luxury: it's a lifeline for making informed decisions.
Public debate oscillates between the catastrophism of "we'll lose our jobs" and the optimism of "increased employment for all." The reality, according to comparative evidence, is much more nuanced: AI will replace tasks, not always entire jobs.This will drive new occupations and demand skills upskilling at an unprecedented pace. We'll break it down with figures, case studies, and recommendations backed by international organizations.
The big numbers: how much employment AI impacts
The International Monetary Fund estimates that AI will reach around 40% of positions worldwideThis impact is spread across jobs that will see automated tasks, functions that will be complemented by technology, and new positions that will emerge. The novelty compared to previous waves of automation is that AI is also making significant inroads into highly skilled occupations (the so-called high-skilled jobs). breaking the stereotype that only "the routine" is replaceable.
The impact will not be the same everywhere. In advanced economies, the IMF anticipates that up to 6 out of 10 jobs They could be affected by AI; in emerging economies it would be around 40%, and in low-income countries, around 26%. The International Labour Organization adds an important nuance: administrative positionsOffices where many women work will suffer significant exposure due to the arrival of generative systems in office tasks.
Spain illustrates another angle of the phenomenon: not only is demand changing, but there is also a shortage of qualified supply. The business ecosystem identified that around 20% of the vacancies in data and AI remained unfilled The last year analyzed by Indesia showed a shortage of specialized professionals. This talent gap coexists with high unemployment rates, which points to mismatches between training and actual needs market.
Regarding net creation and destruction, the numbers vary by source and period, but share a common pattern. World Economic Forum reports discuss tens of millions of jobs replaced and even more created by the reconfiguration of the economy; previous editions quantified the displacement of 75 million compared to 133 million new opportunities (a positive balance of 58 million). Other analyses, such as those by McKinsey, place the figure for 2030 between 20 to 50 million new jobs linked to AI. Beyond the figures, the message is clear: the balance can be favorable if there is large-scale retraining.
What's next: scenarios and business adoption
The OECD emphasizes that, to this day, AI adoption in businesses remains relatively lowHowever, advances in generative AI, falling costs, and a larger pool of trained professionals are placing developed economies on the verge of large-scale expansion. This is no coincidence: generative AI relies heavily on open and high-quality data, which accelerates its spread when access to information is broad and secure.
What role does Spain play in this scenario? The latest data from Indesia anticipates that the national industry will need more than 90.000 specialists in data and AI until 2025. At the same time, the globalization of remote work generates fierce competition: companies around the world are recruiting remote talent with better conditions and more ambitious career plans, so that attract and retain Professionals in Spain need to raise the bar.
Another piece of context: European labor markets have shown good resilience, but Spain is lagging behind. structural unemployment Almost double the EU average, while there is a shortage of qualified professionals in key sectors. With an aging population and strained productivity, the window of opportunity to adjust training and policies is closing. It is not infinite.
There are also operational challenges that hinder full adoption: from infrastructure (stable electricity, Broadband) to gaps in digital skills among staff, technology costs and the need for computerize the economyAll of this suggests that the observable impact will be gradual and will depend on local capacities.
Sectors, occupations and new profiles
Where does AI bite first? OECD surveys of businesses and workers point to the manufacturing and finance among the areas most affected by algorithms and automation. Recent market reports also place retail, hospitality, and transportation as sectors with significant task adjustments, while agriculture, livestock and fishingAssociative activities, extractive industries or construction currently show less aggregate exposure.
The map, however, is not only one of risks. In programming and consulting, scientific and technical services, telecommunications and mediaThis opens up opportunities for new profiles: from experts in natural language processing and prompt engineers...even algorithm auditors, creative specialists capable of managing generative systems or hybrid profiles that combine technical and business skills.
If we look at specific jobs targeted by automation, production and assembly operators appear, professional drivers (autonomous driving is pushing forward), and customer service agents, where conversational assistants are increasingly resolving cases. However, the full story also includes emerging functions such as AI system trainersData analysts and scientists, human-machine team managers, and experts in ethics and governance.
A key principle for orientation: the larger the probability of automating tasks Within a position, the greater the pressure to transform it. This doesn't mean automatic disappearance, but rather redesign of the role, reassignment of functions, and, in many cases, AI support as a co-pilot to gain productivity.
What is being lost and what is being created now
Headlines about AI-related layoffs are multiplying. Major tech companies have cut staff in various areas: around 6.000 positions at Microsoft, more than 8.000 at IBM (mostly in human resources), and some 10.000 cumulative departures at Google since 2023, in addition to cuts at Salesforce, Klarna, Duolingo or Amazon cuts 14.000 jobsIn one striking case, Dukaan's CEO fired the customer service team after demonstrating that an AI system was 85% more efficient.
The media have not been left out either: since 2020 there have been examples of reporters being replaced in certain routine tasks, although some experiments have been partially reversed, as when staff were rehired of attention after discovering the limitations of AI in complex situations. A tech industry layoff tracker counts more than 77.000 affected workers so far this year, about 495 per day (slightly less than the daily average for the previous year), and surveys indicate that 14% of people He claims to have lost a job due to automation.
Are new jobs being created? Yes, although the speed and scale remain to be seen. The World Economic Forum anticipates displacement in some niches and growth in others that are “less automatable” in the short term, with boom in home deliveryConstruction, agriculture, food processing, and nursing are sectors with labor shortages. Furthermore, transitional initiatives are emerging: Ikea offered staff from call centers to retrain as design consultants, and IBM announced its goal of training two million people in AI skills.
The fundamental question is whether businesses, governments, and educational institutions will develop mechanisms to sufficient requalification to absorb those leaving automated tasks and redeploy them to higher value-added roles. Without a roadmap, there is a risk that layoffs will hit harder. entry profiles and vulnerable groups, fueling social divides.
Task-based exposure: a risk map
A comprehensive analysis of more than 400 occupations, based on the standardized international classification, assigns each task a potential automation score between 0 and 1. By combining all the tasks of an occupation, an average and variability are calculated. Based on this, exposure gradients are defined: from the highest (where most tasks have high automatability and little dispersion) to the minimum exposure or almost none.
The findings are interesting: few jobs consist solely of tasks that are fully automatable with current generative AI; almost all occupations include functions that require human interventionFurthermore, exposure is not evenly distributed: female employment appears more concentrated at higher levels (for example, around 5,7% at a high level and 4,7% at a very high level), which calls for gender-focused policies.
By income level, high-income countries accumulate a greater proportion of employment in the gradients described (around the 34%), while in low-income households it barely reaches 11%. And note: we're talking about potential exposure, not actual impact. Full adoption depends on infrastructure, costs and capabilities...among other obstacles. This type of mapping, linked to national microeconomic data, helps to profile social dialogue and public responses precise.
Impact on education and skills that is gaining importance
In the classroom, AI personalizes content, recommends reinforcement activities, and frees up teacher time by automating grading and administrative tasks. Platforms with advanced analytics they adjust the rhythm according to student progress, and conversational assistants offer 24/7 support. Benefits include personalization, global accessibility, continuous assistance, and automation of repetitive tasks.
But it's not all advantages: the data privacyClosing the digital divide that leaves behind those without internet access or devices, and managing reasonable resistance to pedagogical changes. This connects to the world of work: the better we train people in digital skills and critical thinking, the smoother the transition will be labor.
What skills will the economy demand most? STEM (science, technology, engineering, and mathematics) profiles are needed to understand and govern AI, but also other skills. creative, empathetic, and leadership difficult to automate. The trend favors transversal and versatile skills, from communication and customer service to complex problem-solving, project management, and Data ethics.
Updating vocational training and university curricula is a priority: integrating AI, data, and cybersecurity skills, promoting practical projects, and making learning pathways more flexible to accelerate adaptation. In parallel, continuing education within companies must evolve towards Permanent learningwith micro-credentials and agile recycling.
Employment trends and AI in Mexico
Mexico offers an interesting barometer. Since 2021, jobs exposed to AI have grown by nearly a 88%Although there was some slowdown in the job market in 2024, job postings requiring AI skills remained strong. annual compound interest rate of 33,6% between 2021 and 2024, which shows a sustained demand for skilled profiles.
The information and communications sector leads in AI-related job openings, going from approximately one 2,2% to more than 3,6% of the offers. In finance and insurance, manufacturing, health and social work, the share is still less than 1%, indicating slower adoption or different starting points by industry.
Policies and responses: what to do to ensure no one is left behind
The ILO advocates for an orderly and just transition, with a voice for workers. accessible training and social protection adapted to new risks. Without that lever, the benefits could be concentrated in a few countries and companies with better resources. Online, the OECD recommends grounding principles of reliable AI in the workplace, anticipate rights and safety impact assessments, and strengthen oversight and security policies in companies.
Furthermore, governments should encourage new qualifications (and update existing ones), promote training for workers in low qualification and for older workers, integrate AI skills into compulsory and higher education, and promote diversity in the AI workforce. In the labor market, it is advisable to combine active policies (guidance, retraining, and job placement support) with passive policies that protect income without discouraging [the pursuit of new skills/qualifications]. reinstatement to the market.
The institutional framework must facilitate mobility between sectors and occupations to ensure agile reassignment. At the same time, companies would be wise to deploy AI systems that increase the number of people Instead of replacing them outright, invest in continuous learning and design internal transition mechanisms (such as corporate academies) that accelerate the restructuring.
The regulatory response, meanwhile, is progressing at different paces. In the US, discussions are underway about how to hold AI providers accountable for job displacement; in Europe and the UK, transparency obligations are increasing (for example, regarding the use of copyrighted works to train models), with implications for creative industries and new rules of the game for the generated content.
How to prepare yourself professionally
Your individual strategy can make all the difference. First, delve into key technologies (data, automatic learninggenerative tools) and in digital product concepts. Second, cultivate human skills: communication, coordination, leadership, negotiation, and critical thinking. Third, stay up-to-date on trends and regulate your information flow to continuous learning.
Fourth, adopt a collaborative mindset with AI: learn to design human-machine workflows, to assess biases, and explain results Fifth, consider specialized training (from micro-credentials to postgraduate degrees, including MBAs focused on data and AI) to accelerate role changes without losing employability.
There's also a personal risk management component: document your projects, measure impact, build a portfolio showcasing your AI skills, and prioritize organizations that invest in AI. real reskillingnot just in press release promises.
The data paints a picture of a changing world of work: AI will replace tasks, expand others, and trigger new occupations; the impact will be greatest in advanced economies and in administrative roles with a high female presence; some sectors will face intense pressure while others will be more protected; and effective adoption will depend on infrastructure, talent, and policies. Employment doesn't disappear by technological decree, but it will... It reconfigures itself at high speedThe sooner we align education, business, and regulation with that horizon, the better the opportunities will be for workers and companies.