The European Union is attempting to establish itself as a developer of advanced artificial intelligence rather than primarily a regulator and consumer of technology produced elsewhere. An Italian-led consortium developing a large open-source AI model across all 24 official EU languages will provide an important test of whether Europe’s supercomputing infrastructure can support globally competitive systems.

Europe’s efforts to establish greater control over artificial intelligence technology are entering a new phase as the EUROPA consortium, led by Italian company Domyn, advances plans to develop one of the continent’s largest open-source AI models.

The initiative, supported by the European Commission and European high-performance computing infrastructure, aims to create an advanced artificial intelligence system capable of competing with models developed by leading American and Chinese companies.

Domyn chief executive Uljan Sharka told the Financial Times on 29 October that Europe possesses the technological resources and industrial expertise required to compete in advanced AI development.

However, he also identified two obstacles: the need for greater investment and the difficulty of persuading European businesses to adopt domestically developed systems.

These challenges extend beyond the development of a single language model.

They concern whether Europe can establish a commercially sustainable AI industry while reducing dependence on technology providers headquartered outside the EU.

EUROPA and the Frontier AI Grand Challenge

The European Commission selected EUROPA in June 2026 as the winner of its Frontier AI Grand Challenge.

The competition, launched in February, was designed to support the development of a European general-purpose AI model with more than 400 billion parameters.

EUROPA’s planned model is expected to contain approximately 500 billion parameters and support all 24 official EU languages.

The consortium brings together Domyn and Germany’s Fraunhofer research organisation, including the Fraunhofer IAIS and IIS institutes.

Its work will rely on the EuroHPC network of European supercomputers.

Under the competition arrangements, the winning consortium can obtain access to as much as 2.5 per cent of EuroHPC’s total computing capacity for one year.

This allocation is intended to address one of the largest barriers to advanced AI development: the enormous computing resources required to train models at this scale.

The final system is intended to be released under an open-source licence, allowing organisations to inspect, adapt and deploy the technology.

Nevertheless, parameter count alone does not demonstrate that a model is competitive.

Its actual performance will depend on training data, architecture, computational efficiency, evaluation results and the ability to perform reliably across different tasks.

A European Model for European Languages

One of EUROPA’s distinguishing objectives is its coverage of all official EU languages.

The commercial AI market has historically been dominated by models developed primarily for English-language applications, although leading international systems increasingly support multiple languages.

For Europe, linguistic capability has particular relevance to public administration, legal services, education and cross-border business.

A model designed specifically for the European market could potentially support specialised applications involving national legislation, administrative procedures and technical terminology.

However, multilingual availability will need to be assessed against measurable performance standards.

A system capable of producing fluent responses in a particular language is not necessarily equally reliable when interpreting specialised legal or scientific information.

The consortium will therefore need to demonstrate that linguistic coverage translates into practical capabilities rather than remaining principally a technical specification.

Europe’s Supercomputers Become Strategic Assets

The EUROPA project forms part of a wider effort to use European publicly supported computing infrastructure for commercial and scientific AI development.

The EuroHPC Joint Undertaking has invested in systems intended to provide advanced computing resources to researchers, businesses and public institutions.

Additional AI-focused infrastructure is being developed through the EU’s network of AI factories.

These facilities are intended to make specialised computing capacity available to organisations that would otherwise face considerable costs when developing or adapting AI systems.

However, infrastructure ownership does not necessarily provide complete technological independence.

European supercomputing projects continue to rely on internationally sourced semiconductor technology and specialised software.

Domyn itself has described a partnership with the American technology company Nvidia in its previous model-development work.

Consequently, European AI sovereignty is better understood as a question of control over important technological capabilities than complete separation from international supply chains.

Domyn’s Smaller Model Provides an Early Example

On 20 October, Domyn announced the initial release of Domyn Edge, a separate AI model containing four billion parameters.

According to the company’s technical announcement, the system was trained from scratch using approximately 13 trillion tokens on Europe’s Leonardo supercomputer.

Domyn has committed to making the model’s weights, training code, intermediate checkpoints and datasets available under open licences.

This approach is intended to improve reproducibility and allow external developers to examine how the model was constructed.

Domyn Edge is substantially smaller than the planned EUROPA system and should not be confused with it.

Nevertheless, the project provides an example of the company’s approach to developing AI models using European computing infrastructure.

Regulation Is Only Part of the Challenge

The EU has already established an extensive regulatory framework for artificial intelligence through the AI Act.

However, legislation governing the development and deployment of AI does not, by itself, create a competitive domestic technology industry.

Europe faces competition from American companies with substantial private investment, established cloud platforms and access to large commercial markets.

Chinese developers have also expanded their capabilities, including in open-weight AI systems.

EUROPA’s commercial prospects will therefore depend on more than technical performance.

European businesses must be willing to integrate the resulting technology into their operations, while developers will require dependable access to computing infrastructure and support services.

There is also the question of whether European institutions themselves will adopt domestically developed AI systems where they meet relevant security and performance requirements.

Without sufficient demand, even a technically successful model could struggle to establish a sustainable ecosystem.

The Test Will Be Commercial as Well as Technical

Domyn has indicated that it aims to release EUROPA during 2027, although public statements have differed over the precise timetable.

The company’s leadership has discussed a June target, while an earlier technical announcement referred to the second half of 2027.

The immediate challenge is completing development and demonstrating that the model can compete on performance, reliability and cost.

The longer-term challenge concerns adoption.

If European businesses and public institutions continue relying predominantly on foreign technology, access to an advanced European alternative may have only a limited effect on the market.

Conversely, a competitive and widely adopted open-source model could offer organisations greater control over deployment, customisation and sensitive information.

EUROPA is therefore more than a publicly supported technology experiment.

It is a test of whether Europe can convert research institutions, computing infrastructure and industrial expertise into AI products capable of competing in the international market.

The outcome will depend not simply on whether Europe can build an advanced model, but on whether organisations consider it sufficiently capable and economical to use.

A European Commission survey has found widespread adoption of artificial intelligence among smaller businesses, but considerable reluctance to allow automated systems to make decisions about employees. The findings expose a growing divide between technological investment and confidence in workplace automation.

Artificial intelligence is becoming established across European businesses, but almost half of companies remain unwilling to trust automated systems when decisions concern their employees, according to a European Commission survey published on 14th October.

The Eurobarometer findings show that 42 per cent of the small and medium-sized enterprises and mid-cap companies surveyed already use artificial intelligence or algorithmic management systems, although most deployment remains limited or experimental. A further 21 per cent intend to introduce such technologies in the future.

Despite this expanding adoption, 49 per cent of respondents said they did not trust artificial intelligence or automated systems to make staff-related decisions. The figures suggest that companies increasingly distinguish between using technology to improve business processes and delegating managerial responsibility to software.

The Commission’s survey of European businesses was conducted between March and April 2026, involving approximately 12,826 telephone interviews with company leaders across the EU’s 27 member states.

Businesses prefer assistance to automation

The survey provides a more detailed picture of workplace AI than conventional statistics measuring whether companies have adopted the technology.

Among the businesses interviewed, customer service and client relations were the most commonly reported areas of application, accounting for 47 per cent. Monitoring employee performance and productivity accounted for 36 per cent, while logistics, planning and staff scheduling represented 30 per cent.

These functions differ substantially in their consequences for workers and managers.

A system that helps organise deliveries or forecasts customer demand may improve efficiency without directly determining employees’ working conditions. Software that evaluates performance, allocates shifts or recommends staffing changes can influence employment decisions and requires a different level of organisational supervision.

The survey found that 61 per cent of companies considered AI useful in supporting decisions. However, human involvement remained central: 90 per cent of respondents reported that people continued to make decisions, while fully automated decisions accounted for only 2 per cent.

The results undermine the assumption that adopting AI necessarily means transferring control of business operations to autonomous systems.

For many companies, the immediate commercial value lies in processing information more rapidly, identifying patterns and reducing administrative workloads rather than removing human management entirely.

Investment costs remain a substantial barrier

Financial and organisational obstacles continue to limit adoption.

Half of the companies surveyed cited high investment costs, while 49 per cent identified legal uncertainty and 48 per cent pointed to insufficient expertise.

These findings are particularly relevant to small businesses, which frequently lack dedicated technology teams, extensive data infrastructure or the resources to evaluate complex software systems independently.

For such businesses, purchasing AI tools is only one component of the investment. Implementation can also require employee training, changes to internal procedures, data-security assessments and continuing technical support.

The findings add a workplace dimension to concerns raised in EU Today’s earlier examination of Europe’s AI productivity gap, which identified investment, computing capacity and organisational restructuring as obstacles to translating adoption into higher productivity.

A company may purchase sophisticated software without necessarily achieving corresponding improvements in performance. Those improvements depend on whether the technology is integrated into everyday operations and whether employees have the skills to use it effectively.

Brussels prepares further workplace legislation

The survey arrives while the European Commission is considering additional rules on algorithmic management through its proposed Quality Jobs Act.

The initiative is intended to address working conditions, workers’ rights and the increasing use of automated systems in managing employees.

A consultation launched in July asked European employers and trade unions to consider transparency in automated decisions, excessive workplace monitoring and the responsibilities of companies using AI.

The second consultation phase ended in September, allowing social partners to propose negotiated arrangements before the Commission determines its next legislative steps.

The broader regulatory framework is already changing. As EU Today explained when the AI Act entered a new enforcement phase, Brussels has begun enforcing certain obligations for general-purpose AI models, while important requirements for high-risk applications, including employment-related systems, follow a later timetable.

Stand-alone high-risk systems are scheduled to face the relevant obligations from December 2027, with another deadline in August 2028 for high-risk systems embedded in regulated products.

The distinction matters for employers planning investments today. Existing data-protection and employment obligations continue to apply, but some of the AI Act’s more specific compliance requirements remain subject to transition periods.

The Commission’s latest survey indicates that businesses are already exercising caution independently of these future obligations. Their reluctance to automate sensitive staff decisions suggests that accountability, reliability and employee confidence will remain important factors in determining how extensively artificial intelligence is deployed.

For Brussels, the challenge is to establish workable rules for that transition without making implementation prohibitively expensive for smaller companies, which account for much of Europe’s employment and commercial activity.

Just weeks before its scheduled November IPO, Anthropic is facing growing attention over the potential dangers of its technology. The resignation of former Anthropic researcher Jacob Coxon on September 8, who wrote on X that AI could “kill us all by the end of the decade,” preceded the company’s threat report on September 10. According […]

The post AI’s apparent danger is its best selling point appeared first on Asia Times.

The European Union is bringing out the big guns in the face of widespread outrage over the use of Artificial Intelligence (AI) violating people’s privacy and dignity.  

Brussels is considering the possibility of classifying the creation of sexual deepfakes as a prohibited practice under the Artificial Intelligence Act following the scandal involving sexualised images created by Grok AI, the chatbot integrated into Elon Musk’s X platform. 

Grok outrage

Musk’s company xAI – after prolonged international criticism – has introduced new restrictions in mid-January on sexually suggestive AI-generated images in Grok. The move follows criticism that Grok allowed users to digitally replace women’s clothing with bikinis and, in some cases, create sexualised depictions of minors. 

The first images of people being stripped naked without consent (“nudification”) began circulating in the days following the release of the feature, but their spread increased particularly around New Year’s Eve. According to CNN, between January 5 and 6 alone, Grok was used to generate at least 6,700 sexual images. These often involved women or minors. 

“Grok is now offering a ‘spicy mode’ showing explicit sexual content with some output generated with childlike images. This is not spicy. This is illegal. This is appalling,” EU digital affairs spokesman Thomas Regnier told reporters back then.

The European Commission, which acts as the bloc’s digital watchdog, said it would take note of new measures taken by X and would review them. Officials warned that if the steps prove insufficient, the EU will consider using the full scope of its Digital Services Act (DSA).

European Commission Vice-President Henna Virkkunen has stated that the Commission is considering explicitly banning these types of AI-generated sexual images under the AI Act, classifying them as unacceptable risks. 

The prohibition of harmful practices in the field of AI could be relevant to addressing the issue of non-consensual sexual deepfakes and child pornography, said Virkkunen, who is also the responsible commissioner for Tech Sovereignty, Security and Democracy, at this week’s plenary session of the European Parliament in Strasbourg. She also said the DSA mitigated the risk of online sexual material being disseminated without consent. 

She also recalled that the Commission had sent a request for information to X regarding Grok as part of its investigation into the platform under the DSA. 

The platform was asked to preserve all internal documents and data relating to it until the end of the year. “We are now examining the extent to which X may in any case be in breach of the DSA and will not hesitate to take further action if the evidence suggests it,” she said.

The Commission had previously stepped up pressure on X, which was fined 120 million Euro in early December over transparency violations. The EU has insisted it will enforce its rules despite causing backlash from the US administration.

“The DSA is very clear in Europe. All platforms have to get their own house in order, because what they’re generating here is unacceptable, and compliance with EU law is not an option. It’s an obligation,” Regnier said at the height of the scandal in early January. 

Last week, a group of about 50 MEPs had called on the Commission to ban artificial intelligence apps used to create nude images from the EU market. 

Can’t live without X

Despite criticism of X, nearly all senior EU officials continue to post there rather than on European alternatives, according to research by dpa.

European Commission President Ursula von der Leyen and other top officials still do not have official accounts on Mastodon, a Germany-based alternative. Virkkunen  opened an official Mastodon account in January. High-ranking EU politicians are also active on Bluesky, another US-based platform currently gaining traction.

The Commission justifies continued use of X due to its reach: Mastodon has roughly 750,000 monthly users, compared with 100 million on X, according to the companies.

The long legal path towards better safety online

The path towards protection of minors in the EU is long-winded as concerns of privacy, business and protection clash. Several regulations intersect:

• Chat Control

The Commission in 2022 proposed a regulation to require platforms to detect and report images and videos of abuse (child sexual abuse material or CSAM), as well as attempts by predators to contact minors.

Supported by several child protection groups, the plan nicknamed “Chat Control” sparked fierce privacy debates inside the 27-country bloc and led to accusations of mass surveillance. 

The final legislation is expected to be negotiated in early 2026, aiming to bridge the gap between Parliament’s privacy-focused approach and the EU Council’s desire for broad voluntary scanning powers. 

While extending temporary voluntary scanning measures until April 2026 to avoid a legal vacuum, MEPs have expressed urgency for a permanent solution.

• DSA

The European Union uses the Digital Services Act to sanction online platforms by imposing massive fines, requiring immediate operational changes, and – as a last resort – temporarily suspending their services. It can apply fines if platforms violate DSA obligations, fail to comply with interim measures or breach commitments.

It is an EU regulation for a safer online world, requiring platforms to tackle illegal content, protect users, and increase transparency.

• AI Act

The AI Act was adopted in 2024 and is the world’s first and only comprehensive legal framework in the artificial intelligence context. It establishes a risk-based system to regulate AI technologies within the EU, aiming to ensure they are safe, trustworthy, and respect fundamental rights while fostering innovation. 

It bans certain unacceptable AI practices, such as social scoring, and sets rules for areas of high risk for AI use – like in critical infrastructure or employment. It also sets out restrictions on manipulative AI uses such as deepfakes targeting children, among others. 

• Social media bans

France, which is considering banning social media for children under 15, has been testing an age verification app developed by the European Commission since this summer. This tool is one of several methods used to verify the age of internet users, which is a headache for tech giants and authorities alike.

Individual efforts

The Spanish Minister of Youth and Children, Sira Rego, in early January asked the attorney general’s office to investigate whether Grok may be committing crimes related to the dissemination of child sexual abuse material.

Currently, Spain is developing its own law for the protection of minors in digital environments. The law strengthens the framework for protecting personal integrity and privacy against new forms of violation linked to the use of technologies such as artificial intelligence, reaffirming that the best interests of the child must always prevail over any digital business model.

Bulgaria has stepped up efforts to combat online child sexual abuse through international law enforcement cooperation, national prevention campaigns and policy discussions aligned with EU legislation. In 2025, Bulgarian authorities took part in a major international operation that shut down Kidflix, one of the world’s largest platforms for child sexual exploitation, used between 2022 and 2025 by nearly 2 million users.

Romania has legislative mechanisms in place to combat child sexual abuse material through its criminal code, and the authorities are seeking to expand and modernise these rules. 

Since 2025 an important bill on the protection of children online (called the Online Age of Majority Law) has been under parliamentary debate, and Romania is also gradually participating in and implementing EU rules on the prevention and combating of online sexual abuse. The Online Age of Majority Law introduces mandatory age verification and parental consent for minors (under 16) to access online services such as social media, gaming and streaming platforms.

EU candidate country Bosnia and Herzegovina still has no specific law regulating this area. In BiH, criminal liability for the production, distribution and possession of such material is based on criminal laws that cover the sexual exploitation of children, but do not contain explicit provisions on AI-generated or simulated content. 

The EU has put into place a set of complementary tools and measures to protect its citizens – young and old – from harmful practices online, but weak points include challenges in enforcement, algorithmic amplification of harm, inconsistent national implementation, and debates over balancing security with privacy.

This article is an ENR Key Story. The content is based on information published by ENR participating agencies.