Around 38% of people living in Spain regularly use generative artificial intelligence, according to a survey on information and communication technologies in households conducted by the country's National Statistics Institute, or INE. That leaves fewer than four in 10 residents inside the new digital funnel, while 23 million people in the fourth-most-populated EU member country never use these tools, according to the General Union of Workers, or UGT.
The New Inequality
The main use is domestic. Thirty percent of users say they use AI for personal inquiries, nearly double the share who use it regularly at work, at 18% according to the INE and Eurostat, or for educational and training purposes, at 16.2%. The numbers are tidy enough for Brussels-style optimism, but the reality is messier: the technology is spreading fast, unevenly, and on terms set by companies and institutions that keep the gains while the costs are socialized.
UGT says adoption is far from widespread among the general population. In a report translated into English as The Digital Divide and Artificial Intelligence: The New Inequality, the union says the gap is fostering inequality and discrimination in the workforce and undermining social cohesion. The report says the technology has emerged less than four years ago and has spread at an unprecedented pace. ChatGPT, the best-known generative AI brand, reached one billion global users just three-and-a-half years after its launch. By comparison, YouTube took more than twice as long to reach that level of usage, while cellphones required 29 years, more than eight times longer.
Who Gets Left Behind
UGT says the social divide from uneven adoption falls mainly along age, income level and employment status, with older people and low-income earners most affected. The report describes the typical AI user as a young man living in a big city who is employed, earns a good income and has at least finished high school. That’s the profile of access. Everyone else gets the delay, the exclusion, or the lecture about adaptation.
More than half, 55.5%, of people who don't use the technology are women, and the average age of non-users is 56.3 years. The report says there is a 70-percentage-point gap in usage between people aged 16 to 24, 75% of whom use it, and those aged 65 to 74, only 7% of whom do. It says the skills gap affects 59% of people with lower levels of education, while 44% of people in Spain overall have absolutely no knowledge of how to use AI.
UGT also criticizes the gap between these figures and studies commissioned by consulting firms and private entities, which it says claim 51% to 62% of professionals use the technology at work. The union says the INE and Eurostat put the real figure at around 18% of workers, and that there are no rigorous sources supporting claims of widespread workplace adoption. The contrast is blunt. Private reports inflate the story; official figures shrink it.
Workplace Hype, Real-World Lag
The union says the slow uptake at work is not because workers reject the technology, since they use it more in their personal lives, but because companies have failed to integrate these tools into production processes. That’s the corporate bottleneck in plain sight. The machine gets sold as inevitable, then the workplace is left to catch up on someone else’s schedule.
INE data from 2024 shows barely 20% of Spanish companies with more than 10 employees provided information technology training to staff, while for small businesses with fewer than 10 workers, which make up 95% of the business landscape in Spain, the figure drops to 3%. The numbers fit the same old pattern: the biggest firms can afford the training, the smallest are left with scraps, and workers are told the future has already arrived.
UGT says the Spanish government should develop a National AI Skills Plan with training programs open to the general public, establish a right to continuous professional development during working hours and devise a support plan for the unemployed. Those demands point straight at the state’s role in managing inequality after the fact, once the market has already sorted people into users and non-users, winners and leftovers. The report doesn’t describe a neutral technological shift. It describes a hierarchy, built fast, distributed unevenly, and sold as progress to people who won’t all get to use it the same way.