DeepSeek use halted across Belgian federal workplaces

Belgium has ordered all federal government officials to cease using the Chinese AI tool DeepSeek, effective 1 December. All applications linked to the system must be removed from government devices by Monday, according to a letter from Vanessa Matz, Minister of Government Modernisation.

The decision follows an assessment by the Centre for Cybersecurity Belgium, which warned of risks tied to data protection and the handling of sensitive information.

The Belgian government first signalled its concerns in September, after cybersecurity analysts highlighted the tool’s potential to expose transmitted data. Minister Matz stated that the ban is a preventive measure to ensure that federal workplaces remain ‘safe, secure, and exemplary,’ emphasising that vigilance is crucial when dealing with high-risk digital systems, particularly those developed abroad.

Belgium joins a growing list of countries restricting DeepSeek due to security concerns. Taiwan and Australia have already barred the AI system from government devices, citing the risk of data exposure and espionage.

South Korea also banned its use among civil servants after its intelligence service concluded that DeepSeek collected an excessive amount of personal data. Although the company’s app was later restored to app stores, government restrictions remain in place.

Several European governments have taken similar steps. The Czech Republic and the Netherlands have issued limitations, with Dutch civil servants prohibited from installing the chatbot on work devices. Meanwhile, Canada, India, Italy and the United States have imposed their own restrictions, reflecting a broad international push to limit the tool’s reach within public administrations.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!