Statement on export controls for Anthropic Fable 5
The Anthropic debacle highlights the danger of foreign-reliant tech and the urgent need for a sovereign, open, European digital infrastructure.
Following the chaotic release of Anthropic’s Fable 5, the US administration blocked access to the AI model for all non-US citizens through a comprehensive export control directive, effectively shutting down its service. Washington’s unilateral intervention, under its "America First" doctrine is a stark wake-up call, exposing the severe strategic risk of Europe’s digital complacency. With the EU relying on foreign providers for roughly 70% of its cloud infrastructure, we have effectively outsourced our digital sovereignty to foreign executive orders.
Volt Europa calls on European leaders to radically overhaul the Union’s digital infrastructure strategy.
The Fable 5 incident is not an anomaly; it is the most acute manifestation of Europe's structural reliance on non-EU digital infrastructure. The US CLOUD Act of 2018 and FISA Section 702 already grant American authorities extraterritorial access to any data held by US companies, regardless of whether that data is stored on European soil. As EU Commissioner Virkkunen herself acknowledged in the context of the Cloud and AI development Act (CADA) proposal, it remains structurally difficult for American companies to reach the highest European sovereignty levels due to the extraterritorial nature of the US CLOUD Act.
When a single executive order can sever European access to a frontier AI model overnight, and permanently expose European data to foreign government access without European judicial oversight, our systemic dependency becomes undeniable.
The lessons of history: turning Washington’s protectionism into European opportunity
History demonstrates that Washington is repeating a classic strategic blunder. In the 1990s, the US placed encryption software on its military export control list. The results were entirely counterproductive; the tech sector simply moved its key cryptographers and operations to Europe and the UK to bypass the restrictions.
This dynamic illustrates two critical realities. First, the Mobility of Ideas. Innovation moves across borders with total ease, and the internet fundamentally interprets short-sighted laws as damage to be routed around. Second, the sovereignty opportunity. Washington's protectionism is a self-inflicted wound that Europe should leverage as an opening to attract global talent.
Washington's unilateral block on Fable 5 is the modern equivalent of its 1990s blunder. This protectionist overreach creates an immediate commercial and geopolitical opening for the European Union. As the US clamps down on open research, Europe must position itself as the global safe haven for AI researchers, developers, and cryptographers. By offering a stable, predictable, and open regulatory environment, the EU can capture the sovereignty dividend by actively attracting top-tier tech talent and businesses fleeing arbitrary foreign executive mandates.
In contrast, relying on centralised, foreign proprietary tech giants is a systemic security risk for the European Union and its member states. The antidote is therefore structural; when capabilities are baked into open-source software, they cannot be retracted by a foreign executive order overnight.
We should not copy Silicon Valley's past. Instead, we should inherit its future by welcoming the innovators it is currently pushing away.
Demystifying the frontier AI race
True European autonomy cannot be achieved by blindly copying Silicon Valley or racing to accelerate AI development in a vacuum. Frontier AI models like Fable 5 have moved beyond the realm of commercial software into the domain of cyberdefence and critical security infrastructure. When an AI model's underlying capabilities include offensive cybersecurity tools, such as complex vulnerability identification and exploitation capabilities previously confined to nation-state actors, and when these capabilities are protected only by easily circumvented software safeguards, independent technological control becomes a security imperative.
However, the race for the most capable frontier model is largely symbolic. Aside from highly specific security use cases, Mythos-class LLMs like Fable 5 are often only marginally better than their immediate predecessors. Conversely, steep operational costs and resource considerations, such as excessive water usage and severe noise pollution, suggest that the deployment of smaller, specialised models offers a distinct competitive advantage in industrial applications. In some cases desired response times and the strict necessity to keep data local dictate the use of locally deployed models, executed on local hardware, rather than the high-end server farms required by foreign frontier giants.
In this spirit, Volt advocates for companies to design AI architectures focused on sustainability, open standards, and ethical training as best practice over the use of the latest frontier model at all costs.
Actionable policy levers: reforming EU funding and infrastructure
A Europe that runs its frontier models on foreign chips, inside foreign data centres, and under foreign legal jurisdiction has not gained sovereignty. Volt Europa urges the immediate acceleration of the Cloud and AI Development Act (CADA) alongside strict, uniform enforcement of the EU AI Act. We call for extending the spirit of the NIS2 Directive and the Cyber Resilience Act to mandate that essential European infrastructure, from hospitals to energy grids, is built upon open-source, open-standards, and decentralised tech ecosystems that run independently of foreign geopolitical shifts.
To ensure these regulations translate into tangible infrastructure, we endorse EuroStack, the pan-European initiative developed by lawyers, academics, politicians, and tech companies that provides a ready-made blueprint for sovereign, open-source cloud infrastructure. Volt intends to make the principles of EuroStack the operational backbone for implementing the Cloud and AI Development Act.
To back this regulatory blueprint with real infrastructure, Europe must execute concrete fiscal and funding adjustments, specifically by:
increasing funding to reduce foreign tech dependency. The EU must systematically prioritise funding within Horizon Europe and other flagship digital programmes for European participants who build on open, sovereign technology stacks, in line with the CADA's own principle of 'rewarding contributions to EU-based innovation and supply chain resilience’.
funding ideas, not stagnant conglomerates. In line with the recommendations of the IFO Institute and Bocconi University framework, the EU must fundamentally reverse its approach to innovation capital. Currently, the vast majority of EU funding is captured by massive corporations that deliver little actual innovation. Funding must be aggressively redirected toward small and medium-sized enterprises (SMEs) and open-source ecosystems, which serve as the true engines of technological breakthrough.
As Sven Franck, Co-President-elect of Volt Europa, notes, “Europe possesses the talent and technologies to make strides towards technological independence. We also offer the legislative frameworks that foster research and innovation. What is missing is the political will to orient public funding towards European solutions with the clear goal of technological leadership and securing our digital future.”
______________
You and us share the same dream of a united, thriving Europe. It really means a lot to us when you make a donation, and if you would like to help us plan ahead with confidence, we thank you for your monthly contribution 💜