DIGITAL LIFE

Sánchez: AI companies "will never align with the public interest"

Spanish Prime Minister Pedro Sánchez used a meeting with his ministers to comment on Artificial Intelligence—specifically regarding recent developments in the tech world and statements made by leaders of companies in this sector.

At issue are calls from these leaders for greater regulation, which Sánchez interpreted as a possible tactic to secure funding. In the Spanish leader's view, stories about AI models acting on their own initiative might simply be intended to showcase just how advanced the technology has become.

"I believe the risks of uncontrolled Artificial Intelligence are becoming increasingly evident. We see this in numerous news reports about cybersecurity issues stemming from certain applications or models compared to others," Sánchez stated. "But what may also be at play is a strategy by some of these major companies to seek out and attract funding. In short, whatever the case, what is clear is that safety must be an absolute priority in this regard as well."

Regarding safety, Sánchez emphasized the need to "strengthen cyber protection" and to "increase oversight of the most aggressive models."

Sánchez also addressed the topic of regulation, stating that "self-regulation does not work"—a lesson taught by history.

The Spanish head of government recalled William Vanderbilt, a key figure in the expansion of US railways in the late 19th century, who famously declared: "The public be damned. I work only for my shareholders." “I think that phrase symbolizes many things, but what I would like to highlight is that it exemplifies, perhaps better than almost any other, the spirit of that era: the arrogance of a class that believed itself to be above everything and everyone. The only regulation it accepted was that which might originate from within itself—never from public authorities, whom, incidentally, this statement made clear they held in contempt,” noted Sánchez.

Drawing a parallel with the present day, Sánchez argued that such regulation must come from the state itself; one cannot expect companies to act against their own interests.

“The incentives of the few who control cutting-edge industries will never align with the public interest, no matter how much they try to convince us otherwise,” declared Sánchez.

What is the IA360 Plan...The IA360 Plan seeks to address artificial intelligence from all its angles: economic, labor, educational, technological, and regulatory. Hence the reference to the “360 degrees” of its name.

The strategy is structured around four major objectives: to reach a national agreement on the implementation of artificial intelligence, to strengthen Spain's own technological capacity, to make AI an economic engine and a talent creator, as well as to improve regulation, supervision, and cybersecurity.

The central idea is that Spain should not be limited to consuming tools developed by large foreign companies, but should have infrastructure, its own models, trained workers, and rules to control how this technology is used.

A “new social contract” for artificial intelligence...Sánchez has announced that the Government will summon social agents next month —mainly business organizations and unions— to advance in a “major national agreement” on artificial intelligence. The objective is to anticipate the changes that AI will cause in the labor market. Technology can automate tasks, modify professions, increase productivity, and create new jobs, but it can also displace workers, increase surveillance in companies, or introduce algorithms in decisions such as hiring, performance evaluation, or layoffs.

The agreement aims to provide security to workers and establish shared rules on issues such as: the adaptation of jobs, the training and reskilling of employees, the use of algorithms within companies, the protection of labor rights, the distribution of productivity gains, and the transition of the most exposed sectors and professions.

It is not, for now, a closed labor reform nor a new concrete law. The Government proposes to open a dialogue process with employers and unions to build what Sánchez has defined as a “new social contract” for artificial intelligence.

The president has defended that this consensus must also incorporate workers, political parties, public administrations, and other sectors of society.

An observatory will study the impact of AI on employment...The IA360 Plan includes the creation of a permanent observatory on artificial intelligence and the labor market. Its function will be to analyze how the implementation of this technology evolves and what effects it produces on employment.

This body must help determine what tasks are being automated, what professions are changing, where jobs are being destroyed or created, and what new skills workers need. The purpose is to have updated information to adopt measures before technological changes generate imbalances that are difficult to correct.

Sectoral tables will also be established to study the necessary transitions in areas such as education, the financial system, the labor market, or culture.

In this last case, the Government foresees a specific table with representatives from the cultural sector to analyze the effect of artificial intelligence on creation, authors' rights, and the economic value of content.

A gigafactory to reinforce technological autonomy...The second axis of the IA360 Plan consists of reinforcing Spain's “technological muscle.” One of its main projects will be the candidacy to host one of the seven gigafactories of artificial intelligence planned by the European Union.

An AI gigafactory is a large infrastructure equipped with thousands of advanced processors, enormous storage capacities, and high-speed networks. It allows training and executing large-scale artificial intelligence models without completely relying on the technological centers of the United States or China.

The plan also includes the development, together with the Barcelona Supercomputing Center, of models intended for specific applications in areas such as health, energy, the fight against the climate emergency, and the improvement of public services.

Sánchez has added that the new data centers must meet demanding environmental and energy standards, respect natural resources, protect data autonomy, and generate benefits for the territories where they are installed.

The goal: for half of SMEs to use generative AI by 2030...The third axis aims to extend the use of artificial intelligence beyond large companies. The goal set by the Government is that half of the small and medium-sized Spanish enterprises integrate generative AI into their processes before 2030.

In practice, this may mean using it to automate administrative tasks, serve customers, analyze information, prepare budgets, optimize production, or design new products. The challenge is to avoid a gap between large companies, which have the resources to adopt these technologies, and small businesses, which may face more economic and technical difficulties.

The Government links this objective to training. The IA360 Plan foresees adapting the curricula of Secondary Education and Vocational Training so that young people learn to use AI responsibly and productively, without these tools harming their learning.

More control over the most advanced artificial intelligence models...The fourth block of the plan focuses on governance and security. Sánchez has announced the strengthening of the Spanish “cybersecurity shield” and greater oversight of the so-called frontier models.

These are the most powerful and advanced artificial intelligence systems, capable of performing complex tasks and which, precisely because of their capability, may present greater risks if used to launch cyberattacks, spread misinformation, or make decisions without sufficient controls.

The president has called for artificial intelligence with oversight and traceability. That is, systems whose operation can be monitored and whose decisions can be reconstructed or explained, especially when used in areas where public interest is at stake.

AI companies may never fully align with the public interest primarily because their core financial incentives conflict with societal well-being. While public interest prioritizes safety, equity, job stability, and truth, commercial AI development is driven by a market structure that rewards speed, data collection, and profit maximization.This inherent tension is often described by tech critics, economists, and ethics researchers through several key arguments:

1. The profit incentive vs. public good:

Shareholder Primacy: As heavily funded corporations or venture-backed startups, AI companies have a legal and financial duty to maximize returns for investors. When safety testing or ethical guardrails slow down product launches, profit incentives usually win.

The "Move Fast and Break Things" Culture: Silicon Valley relies on being first to market to capture monopoly-like market share. This race discourages the slow, deliberate auditing required to ensure a technology does not harm the public.

2. Data and privacy exploitation:

Surveillance Capitalism: The business model of modern AI relies on scraping massive amounts of data, often without explicit consent, compensation, or regard for individual privacy.

Value Extraction: AI systems train on the collective knowledge, art, and writing of the public to create commercial products that could ultimately automate away the very livelihoods of the people who created that data.

3. Asymmetry of power and accountability:

Regulatory Capture: Tech giants have massive lobbying budgets. They often influence legislation to protect their own market positions rather than to protect consumers, effectively writing the rules of their own oversight

Lack of Democratic Input: Decisions about what AI models are built, what biases they contain, and how they deploy automated systems are made behind closed doors by a small group of tech executives, rather than through democratic or public processes

4. Externalization of costs:

AI companies capture the massive financial profits of automation while shifting the negative consequences onto society:

Misinformation: Maximizing user engagement often means algorithmic feeds promote sensationalist, AI-generated disinformation, damaging democratic discourse.

Environmental Impact: Running massive data centers requires immense amounts of water and energy, contributing heavily to carbon emissions while the public bears the climate costs.

Economic Disruption: Rapid job displacement affects workers and state social safety nets, while the wealth generated by AI concentrates into fewer hands.

Miguel Dias