Spain unveils new AI plan as Sánchez warns unregulated technology could have ‘devastating’ consequences

Speaking at the Moncloa in Madrid on Monday, Sánchez rejected the idea that the rapidly expanding AI industry could be left to police itself, arguing that governments and society must establish rules governing a technology that is developing at extraordinary speed.

‘If artificial intelligence is concentrated among a few companies and a handful of countries, its impact on shared prosperity, and therefore on the idea of democracy, will be very negative, if not devastating,’ he said.

‘We must act urgently, because unregulated, opaque artificial intelligence running wild … and in very few hands is the perfect recipe for disaster.’

The comments came as Sánchez closed an event presenting the Spanish government’s Plan IA360, a roadmap for the responsible deployment of artificial intelligence over the coming 12 months.

The plan combines investment in computing infrastructure and new AI models with measures covering cybersecurity, education, employment, data centres and regulation. The government also wants to begin negotiations on what Sánchez described as a new ‘Social Contract for Artificial Intelligence’, intended to protect people as the technology changes both workplaces and everyday life.

‘Self-regulation does not work’

Sánchez used some of his strongest language to challenge the influence of the handful of technology companies driving the latest advances in AI.

‘Self-regulation does not work,’ he said, arguing that the commercial incentives of those controlling leading industries would not necessarily coincide with the wider public interest.

He referred to an interview in which one of the industry’s leading figures had essentially asked the public to trust technology companies, although he did not identify the person concerned. Sánchez said his response was effectively: no.

The prime minister, who has previously criticised what he calls ‘techno-oligarchs’, compared today’s debate with the enormous corporate power accumulated during the United States’ Gilded Age in the late 19th century. ALSO READ: EU proposes social media ban for under-13s as Sánchez says Europe is standing up to ‘techno-oligarchs’.

He began his remarks by recalling railway magnate William Vanderbilt and the famous remark attributed to him that he was concerned with his shareholders rather than the public. Sánchez argued that, 150 years later, some business leaders were again presenting government regulation as an obstacle to innovation.

He said powerful companies could not be permitted to determine the rules governing themselves and argued that AI required clear rules established and supervised collectively.

Without collective human control, Sánchez said, artificial intelligence risked becoming ‘chaos’ rather than intelligence.

Growing alarm within the AI industry

His intervention comes amid an increasingly intense international debate about whether the development of the most powerful AI systems is moving too quickly.

Prominent figures in the industry have themselves raised concerns about the potential dangers of increasingly capable systems, including the possibility that future AI could improve its own capabilities and eventually become difficult for humans to control. 

Recent warnings have been supported by leading figures including Dario Amodei, Sam Altman, Elon Musk and Demis Hassabis, while former employees of major AI companies have also spoken publicly about risks posed by the technology. 

Sánchez, however, expressed some scepticism about the motivation behind at least some of the industry’s increasingly dramatic warnings.

Referring to recent statements by leading AI figures, he said it was difficult to know exactly how much truth lay behind them, while suggesting that they could also form part of efforts by some companies to attract financing. ALSO READ: Calls for an AI slowdown: conscience, politics, or cold feet?

Whatever the motives, he argued, the warnings reinforced the need for governments to treat AI security as a priority.

Spain therefore intends to strengthen its cybersecurity defences and increase supervision of the most advanced or potentially aggressive frontier AI models. 

Gigafactory and Spanish-developed AI models

At the same time, the government insists that tighter oversight should not mean abandoning the economic opportunities offered by artificial intelligence.

Plan IA360 includes proposals for an AI gigafactory to increase Spain’s computing capacity and the development, together with the Barcelona Supercomputing Center, of new AI models intended for areas including healthcare, energy and tackling the climate emergency. ALSO READ: Spain seeks leading role in European AI push with €719m investment.

The government also plans an AI voucher scheme designed to help small and medium-sized companies and self-employed workers adopt the technology.

Its target is for more than half of Spanish companies to have incorporated AI into their operations by 2030.

Education will also form part of the strategy, with changes planned to secondary-school and vocational-training curricula so that young people learn to use AI responsibly, while attempting to prevent the technology from undermining the learning process itself.

Sánchez said AI could either help close social, economic and geographical divides or make them wider. Similarly, it could create new employment or eliminate more jobs than it generates.

Data centres face scrutiny over water and energy

The environmental cost of the AI boom is another area the Spanish government intends to address.

The huge computing infrastructure required for increasingly powerful AI systems has led to rapidly expanding demand for data centres, which in turn consume substantial amounts of electricity and, in some cases, water.

Sánchez said new facilities in Spain would be expected to meet high environmental and energy standards, protect the country’s autonomy over the data stored in them and provide tangible benefits to the communities where they are located.

‘We must develop this technology not behind citizens’ backs … but in a way that people see it as an opportunity rather than a tool for depleting their scarce resources,’ he said.

The government wants communities to be able to judge proposed data centres partly according to their impact on resources such as electricity and water.

A new ‘Social Contract’ for AI

Perhaps the most politically significant element of the strategy is Sánchez’s proposal for a national agreement involving government, employers, trade unions and other parts of civil society.

The prime minister said social partners would be called together from next month to begin work on a new Social Contract for Artificial Intelligence.

Its purpose would be to provide greater security for people confronted with changes to employment and everyday life brought about by AI.

Sánchez also called for technology companies to face legal and financial responsibility for damage caused by their algorithms, pointing to areas including deepfakes, harm to the mental health of children and young people, and the theft of intellectual and industrial property. ALSO READ: Spanish government approves draft law to crack down on AI deepfakes and image misuse. 

His government argues that a completely new international institution to oversee AI is unnecessary. Instead, Sánchez wants existing mechanisms strengthened, including the European Union’s AI Office and the United Nations’ scientific work on artificial intelligence.

Spain intends to work with countries including Canada and Japan, as well as other EU member states, to push for greater international coordination, arguing that no individual country can effectively supervise a technology that operates across national borders.

Spain seeks a leading role in AI regulation

Spain has already attempted to position itself near the forefront of international AI governance.

It became the first country to establish a dedicated artificial intelligence supervisory authority, the Spanish Agency for the Supervision of Artificial Intelligence (AESIA), which is based in A Coruña.

Sánchez also highlighted the agreement reached on the EU’s Artificial Intelligence Act during Spain’s presidency of the Council of the European Union, describing the legislation as an important first step towards regulating the technology at a supranational level.

Spain’s own national AI legislation, however, has yet to complete its passage through parliament, meaning that EU rules remain central to the country’s binding regulatory framework.

The prime minister argued that the fundamental question was no longer whether AI should continue to develop, but under what conditions that development should take place.

Technology, he said, could increase productivity, create opportunities and make employees more valuable rather than more dispensable — but none of those outcomes was automatic.

For Sánchez, the speed of development makes the issue particularly urgent: AI has achieved in only a few years advances that previous general-purpose technologies took decades to accomplish.

The future impact of the technology, he argued, therefore remains a political and social choice rather than something that should be determined exclusively by the companies developing it.

‘Progress in technology does not guarantee social progress,’ Sánchez said in closing, insisting that governments, companies and citizens must collectively determine how artificial intelligence is allowed to reshape society.