Germany plans AI safety institute to assess risks of advanced models
Germany has announced plans to establish a national AI safety institute that will evaluate the capabilities and risks of advanced AI systems, with an initial focus on cybersecurity implications.
The German government has announced plans to create an AI safety institute tasked with assessing the performance, capabilities, and risks of advanced AI models.
The decision follows discussions within Germany’s National Security Council on the cybersecurity implications of increasingly capable AI systems. According to government officials, the new institute will analyse advanced models, examine potential risks associated with their deployment, and strengthen information-sharing with similar organisations in other countries.
German authorities said the initiative aims to foster a better understanding of how advanced AI systems could affect national security and cybersecurity. The institute is also expected to contribute to international cooperation on AI governance and technical standards.
According to a government spokesperson, one objective will be to work with international partners to develop harmonised approaches to AI safety and responsible use.
The announcement comes as Germany continues to strengthen its cybersecurity policies amid growing concerns about cyber threats. The Interior Ministry reported approximately 334,000 cybercrime cases last year, with around two-thirds originating from abroad or from unidentified sources.
The planned institute places Germany among a growing group of countries establishing dedicated technical bodies to evaluate advanced AI systems. Similar organisations have been launched or proposed in jurisdictions including the United Kingdom, the United States, Japan, Singapore, and South Korea, reflecting increasing government interest in developing independent expertise on AI safety, security, and risk assessment.
Details regarding the institute’s structure, mandate, and timeline have not yet been announced.
