Overview
- The Federal Government's National Security Council decided on June 8–9 to establish a national AI Security Institute to consolidate work on analysing the capabilities and risks of advanced AI models.
- Officials say the institute's mandate will include technical testing of AI systems, advising the government and exchanging intelligence with foreign counterparts to push for harmonized international standards.
- The move re‑bundles functions long handled by the Federal Office for Information Security (BSI) but the council has not yet released details on the institute's legal form, location, staffing or implementation timeline.
- At the same meeting the council ordered a review of Sicherstellungs‑ und Vorsorgegesetze to update rules that secure vital supplies and critical infrastructure for civilian and military needs in crises.
- Germany's step follows other countries that already run specialised AI security bodies and could speed government access to test results and advice, which may affect how quickly new models are approved for use and how regulators enforce the EU AI Act.