Europe's AI translation industry faces reputational risks by partnering with US firms, according to industry figures, after leading startup DeepL decided to collaborate with Amazon Web Services (AWS). The move has sparked alarm over Silicon Valley's monopoly on digital infrastructure.
DeepL's Partnership with AWS
DeepL, a Cologne-based online translator that consistently outperforms Google Translate in accuracy assessments, recently informed paying subscribers that it would no longer process data exclusively on its own servers and would partner with AWS for infrastructure. The company emphasized that AWS would not have access to customer data for viewing or training algorithms, describing the partnership as vital for international scaling.
A DeepL spokesperson stated: "DeepL remains the data processor. We have added AWS as a sub-processor to provide necessary infrastructure for global scale. AWS will not control or access customer data in any usable form. Customer data is encrypted in transit and at rest, and we do not use customer data from paid services to train our AI models."
User Concerns Over Data Sovereignty
Despite these assurances, some long-time business customers have decided to cancel subscriptions. Jörg Weishaupt, CEO of Malogica Group, expressed discomfort uploading confidential contracts or strategy papers, citing concerns about US legislation such as the Patriot Act and Cloud Act, which allow US authorities to request data from cloud providers. Weishaupt noted a growing movement toward data sovereignty in Europe, stating: "We are all trying to get out of the lock-in with the Americans."
DeepL offers a data residency option to guarantee European data remains within the continent, but some question its reliability. The geopolitical climate, including clashes between the Trump administration and EU over tech regulation, has heightened sensitivity around data independence. European Commission President Ursula von der Leyen recently described taking control over technologies as "Europe's independence moment."
Industry Perspectives
Marco Trombetti, CEO of DeepL competitor Translated, warned against relying on US infrastructure, calling it a "big, big error." He argued that European AI translation companies thrive due to their multilingual market and timezone advantages, and that US chip export restrictions create an uneven playing field. "US companies having priority access to chips creates a strong incentive to relocate to the United States," he said, adding that such restrictions motivate China and the EU to build alternatives.
Leevi Saari, a researcher at the University of Amsterdam and AI Now Institute, noted that scaling services requires significant capacity, making AWS tempting despite outages. He highlighted the challenge of datacentre latency, which AWS addresses through a global network of datacentres and subsea cables. "Currently, the gravitational forces of the AI industry are such that startups will end up being pulled towards the US," Saari said. "How can Europe create its own AI gravity well? That's the trillion euro question."
Building a European digital infrastructure remains a major challenge, but the debate underscores the tension between leveraging US cloud services and maintaining strategic autonomy in AI.
