Since the end of the 20th century, the world has witnessed a tendency towards the reduction of income inequalities between countries, driven by the accelerated growth of emerging economies such as China and India and for the effects of globalization, which has facilitated the transfer of capital and technology from rich countries to developing countries. However, this international convergence contrasts with a growing income inequality within the countries themselves, a phenomenon that is accentuated since the 1980s, especially in the United States, although it is also present in Europe. One of the main engines of this internal inequality has been automation. By replacing routine tasks through machines, labor demand is reconfigured, as it decreases in low qualification occupations and increases in those that require advanced technical and cognitive skills. This phenomenon generates an unequal impact that penalizes those who occupy easily automatized jobs, often people with less training. In addition, automation tends to increase the weight of capital over work in the distribution of income, which further reinforces the concentration of income. The emergence of artificial intelligence (AI) intensifies these dynamics. Unlike previous automation stages focused on physical or repetitive tasks, AI can assume intellectual, analytical and creative tasks. Thanks to the massive use of data and continuous learning algorithms, AI is able to classify, predict and generate useful patterns for decision making in sectors as diverse as logistics, finance or human resources. This expansion of its scope threatens to deepen inequality by concentrating even more economic power in the hands of those who control these technologies, while polarizing the labor market between those who dominate their use and those who are lagging behind. In this context, the generative the AI represents a new border. This AI variant not only interprets and predicts, but also creates content (texts, images, codes) from human instructions. Although its concentrator potential is undeniable, it can also become, if it is properly oriented, a democratizing tool. Daron Acemoglu, Nobel Prize in Economics, emphasizes that generative artificial intelligence has great potential to reduce inequalities if it develops and applies following a complementary approach to human work. Instead of focusing solely on automation – which tends to displace workers and accentuate salary gaps – can be used to expand the abilities of a wide range of workers, including those without university degrees. Well designed tools can facilitate access to knowledge, improve decision making, accelerate learning and allow people with less training to perform more qualified tasks. This would not only raise productivity, but also improve the quality of employment and expand work opportunities. For this, it is essential that capable tools be designed to solve complex and non -routine problems, (2) provide useful information at the right time to facilitate decisions, (3) compensate technical or linguistic lacks in vulnerable groups (such as manual workers or migrants), and (4) Promote professional training and recycling processes, especially in essential sectors such as education, health or technical trades. They illustrate how this vision can materialize: the first is the American program Modern Craft Workers, which combines investments in infrastructure with professional training. Through the use of generative, workers without higher studies receive real -time support to perform complex tasks in sectors such as construction, repair, material design or transport. This technology acts as an «expert assistant» that expands the autonomy of manual workers and improves their employability. The second example focuses on the educational field. The generative AI allows to offer custom tutorials to students who require reinforcement to a very accessible cost. These tools can reduce learning barriers for students at a disadvantage, facilitating their access to key competences for a future quality labor insertion. In summary, although automation and artificial intelligence have so far been catalysts of a growing inequality, there is a real potential to reverse this trend if public policies and inclusive technological strategies are adopted. The key is not to leave the evolution of AI exclusively in the hands of the market, but to guide its development towards the common good, so that instead of replacing people, empowers them. The generative AI has the real power to do so, but this requires a redirection of technological innovation, as well as corporate norms, towards priorities that put people in the center of the objectives of this development. It is a non -trivial challenge, no doubt, but not less necessary.
