> ## Content Index
> Fetch the complete content index at: https://www.nupath.eu/llms.txt
> Use this file to discover other available public pages before exploring further.

# EU urged to preserve open web access for AI sovereignty
- URL: https://www.nupath.eu/eu-urged-to-preserve-open-web-access-for-ai-sovereignty/
- Published: 2026-11-18T06:26:21.000Z
- Updated: 2026-11-18T06:26:21.000Z
- Author: Christian Triantafillou Schade
- Tags: News, #source_EU-Startups

An opinion piece argues Europe’s AI sovereignty ambitions depend on unrestricted access to publicly available web data for training models. It warns that emerging licensing models and blanket access blocks by private companies risk excluding EU startups from innovation and competition. The analysis cites a 15-fold increase in AI agents and research showing one-third of leading commerce sites restrict automated access. **(EU-STARTUPS)**

[Opinion: Maintaining the open web is central to Europe’s AI sovereignty ambitionsEurope’s ambitions for AI sovereignty depend on giving its startups the resources to succeed, innovate and compete with established companies. As it stands, the EU remains heavily reliant on digital technologies developed and managed from elsewhere, raising questions about its ability to control the systems underpinning its economy. In June this year, the European Commission unveiled its European Technological Sovereignty Package. This includes the Cloud and AI Development Act, which specifically aims to support the research, development and innovation of next-generation AI technologies, including frontier AI models. The Tech Sovereignty Package also aims to fast-track the implementation of the European Commission’s existing AI Continent Action Plan. This encompasses many areas that are central to the EU’s AI sovereignty ambitions, including physical infrastructure, talent development, and regulatory simplification. One of the most notable of these is access to large volumes of high-quality data, which the plan specifically highlights as a requirement for bolstering AI innovation. For European AI startups, publicly available web data is a valuable resource for training models and building services that rely on current information. Yet emerging licensing models that treat public data as exclusive, regulatory uncertainties and restrictions on automated access risk putting that resource beyond the reach of smaller developers. Maintaining responsible access to the open web is therefore central to enabling businesses of all sizes to contribute to the EU’s AI sovereignty ambitions. AI sovereignty depends on an open web The internet itself was built on the principle that information can be published and accessed widely, without any one company controlling access to it. Within a few years of launching the World Wide Web, CERN placed the software in the public domain for anyone to use, modify and distribute. This openness allowed businesses, researchers and developers to build on shared foundations without seeking permission from a single technology provider. This principle underpinned, and continues to support, both the development and functioning of frontier AI systems. Varied public web data can help reduce gaps and biases in training data, helping AI models to deliver accurate, unbiased outputs. The rise of agentic AI has further increased the importance of access to public web data. It is being deployed across a growing range of services, from personal shopping and travel assistants to business research and IT support. Microsoft’s Work Trend annual index report found a 15-fold increase in active agents across its Microsoft 365 ecosystem between March 2025 and March 2026\. For agents carrying out tasks on the web, however, restrictions on access to current information can limit their effectiveness. For example, an agent comparing suppliers or booking travel needs to check changing prices, availability and terms before acting. In October, research found that roughly a third of the several hundred leading US commerce, travel and reservation websites it tested blocked Meta’s Muse agent, while another third challenged or restricted its access. The biggest problem here is not for Meta, who have been negotiating leverage and large financial resources for legal disputes, but many smaller agent developers do not. This limits their ability to innovate, as their agents have less access to up-to-date information. Consumers also lose out, with fewer competing services to choose from and agents that may be unable to compare the full range of products, prices and offers available. Such arbitrary restrictions threaten the EU’s AI ambitions as its AI startups are excluded from the market by the decisions of competing private companies. They also raise questions about consumer sovereignty in the free market and how it may be limited. While regulators discuss opt-outs, blanket blocks are implemented In September 2026, the European Commission launched a consultation on whether the EU’s copyright framework needs additional measures, including those addressing the use of copyright-protected content in AI. The resulting policy decisions could have significant implications far beyond AI. A key risk is that overly broad measures have unintended consequences as data resources previously used for innovation and value creation become virtually useless for society. They become exclusive to the biggest market players. For European startups, this translates directly into a competitive disadvantage against companies operating under more proportionate regulatory regimes. While the EU is still considering the scope of copyright holders’ opt-outs and how they should work, private companies are already implementing their own mechanisms, waiting neither for regulation nor for the website owner to actually make the choice. ​​If web hosts and platforms get to make such decisions instead of the users that created the content or the websites that published it, copyright will slowly lose its meaning. This will not be because of AI companies accessing content but because of the lack of control publishers will have over who can access their content. Both AI companies and publishers are seeking clarity over who can access content, for what purposes and on whose terms. These questions should be resolved through clear rules that give rights holders meaningful control, preserve lawful access to public web data and hold those who misuse it accountable. The web must remain open. This means lawful information published openly on the web should remain accessible. Personal, private information should remain protected, while the collection and use of personal data must comply with data-protection laws. And this is where the issue lies. We need to clearly distinguish what information is public and what information is deemed private. Beyond these regulatory questions, the risk for EU startups is the creation of a pay-to-play economy that favours large technology companies and well-financed incumbents with negotiated access to public data. This raises the cost of innovation and restricts European startups’ ability to catch up and compete, undermining the EU’s AI sovereignty ambitions. Enabling the EU’s AI sovereignty ambitions Creating a commercial and regulatory environment that supports innovation and global competitiveness is central to developing the EU’s sovereign AI ecosystem. Existing regulations and ongoing policy initiatives such as the AI Act provide an important foundation, but further work is needed to clarify compliance requirements and address barriers that favour well-financed incumbents. Access to the same publicly available information should not depend solely on a company’s financial resources or its ability to negotiate preferential deals. The EU needs to balance legitimate access to public web data with the protection of copyright and platforms’ commercial interests. These objectives can coexist, but developers and publishers need clearer rules about what information can be collected, for which purposes and under what conditions. Ultimately, the EU’s AI sovereignty ambitions depend not only on investment in infrastructure and talent but also on European companies’ ability to access and use diverse, up-to-date data. The post Opinion: Maintaining the open web is central to Europe’s AI sovereignty ambitions appeared first on EU-Startups.![](https://www.google.com/s2/favicons?domain=https://www.eu-startups.com/2026/11/opinion-maintaining-the-open-web-is-central-to-europes-ai-sovereignty-ambitions/&sz=64)EU-Startups![](https://www.eu-startups.com/wp-content/uploads/2026/11/Opinion_-Maintaining-the-open-web-is-central-to-Europes-AI-sovereignty-ambitions-2.png)](https://www.eu-startups.com/2026/11/opinion-maintaining-the-open-web-is-central-to-europes-ai-sovereignty-ambitions/?ref=nupath.eu)