Germany - AI Action Plan 2023

Overview

The BMBF AI Action Plan 2023 ("Aktionsplan Künstliche Intelligenz 2023") was published as the Federal Ministry of Education and Research's targeted implementation update to Germany's National AI Strategy. Formally released on 7 November 2023, the Action Plan frames the emergence of generative AI (the "ChatGPT moment") as a catalyst demanding accelerated, coordinated action across research, infrastructure, skills and application domains. The document sets three strategic objectives (strengthening foundations, deepening European cooperation, and targeting application value) and articulates eleven priority action fields with concrete measures and funding commitments. The full text is published by the ministry and is available as a downloadable PDF on the National AI Strategy portal. For the official PDF see BMBF - Aktionsplan Künstliche Intelligenz 2023 (PDF).

Definitions

The Action Plan adopts the meaning of "Artificial Intelligence" consistent with the National AI Strategy: an operational, application-oriented notion of AI that emphasizes methods (machine learning, pattern recognition, robotics, multimodal interaction) used to solve concrete problems rather than speculative notions of "strong" AI. The plan distinguishes foundational research (methods, architectures, model development), applied research (domain-specific use cases), and enabling infrastructures (data repositories, high-performance computing, testbeds). It underscores "trustworthy AI" principles — human-centred, transparent, robust — without replacing statutory legal definitions being developed at EU level (e.g., the EU AI Act).

Governance and Institutional Framework

The Action Plan is a ministry-level implementation document: the BMBF is the lead federal sponsor and coordinator for the measures described. The plan mobilises existing research actors (universities, competence centres, the DFKI, Helmholtz, Max Planck, Fraunhofer, and university hospitals) and funding instruments, while committing to new targeted calls and an estimated BMBF funding envelope of more than EUR 1.6 billion within the legislative period to support the measures. It foresees stronger networking of competence centres, expansion of professorships, and the establishment of testbeds and secure research environments. Governance is implementation-focused rather than regulatory: the document asks for inter-ministerial cooperation, closer EU-level coordination, and multi-stakeholder engagement (industry, civil society, research). The Action Plan is published on the government AI strategy portal; see the official hosting and publisher details at KI-Strategie Deutschland (publisher information).

Key Focus Areas

The eleven priority fields set out in the Action Plan concentrate policy and funding activity. They include: (1) strengthening the research base (new model and method funding, neuro-inspired AI networks); (2) setting a research agenda for novel AI paradigms; (3) expanding AI infrastructure and compute accessibility for research and SMEs; (4) a competence offensive (professorships, training, lifelong learning); (5) boosting transfer into industry and SMEs and facilitating uptake by public administration; (6) promoting AI in healthcare (data infrastructures, NUM linkages, Digital Progress Hubs Health, clinical and care applications); (7) unlocking scientific and societal uses of AI; (8) researching and designing AI tools for the education system; (9) intensifying European and international partnership and alignment; (10) advancing societal dialogue and multidisciplinary research (ethics, social sciences); and (11) creating an innovation-friendly, agile regulatory approach oriented to European rule-making. Many measures are technical (compute and data), organisational (networks, professorships), and programmematic (funding lines, testbeds) in character.

Implementation Framework

Implementation uses three main instruments: targeted funding calls and instruments managed by the BMBF; stronger coordination of existing programmes and competence centres; and strategic European engagement to leverage shared infrastructures. The plan states that approximately 50 active measures will be reprioritised and at least 20 new initiatives launched. Concrete implementation steps include the establishment of secure testing environments for AI security research, new funding for machine-learning model development, the enlargement of university and applied research posts, and measures to make HPC and cloud-based compute resources more accessible to researchers and SMEs. Implementation responsibility rests with the BMBF for the measures listed; many require cooperation with other ministries, Länder, research institutions and EU partners. Further operational details and funding instruments are to be published in separate calls and programme notices by the ministry.

Monitoring and Evaluation

The Action Plan commits to monitoring progress through programme-level evaluation and milestone tracking tied to funding calls. The BMBF intends to report on implementation of the measures and to use evaluation findings to re-calibrate funding and activities. Monitoring covers outputs (new professorships, centres, funded projects), infrastructure metrics (compute capacity, data availability), and uptake indicators (SME use of AI, transfer projects). The plan also foresees periodic exchange with European partners and an expectation of reporting within the broader National AI Strategy monitoring structures.

Penalties, Liability, and Appeals

As a policy and funding plan, the Action Plan does not create novel criminal or administrative penalties. It outlines expectations, funding conditions and programme rules, which will be implemented through normal grant conditions and public funding obligations; non-compliance with grant terms can trigger standard grant recovery and administrative remedies under public funding rules. For legal liability in AI deployment, the plan points to statutory frameworks (including EU-level instruments such as the EU AI Act) as the source of enforceable obligations; the Action Plan states that agile, innovation-friendly regulation will be pursued in coordination with other ministries and at EU level.

Relationship to Other Instruments

The Action Plan directly updates the BMBF contribution to the National AI Strategy (2018) and its 2020 update. It complements and works alongside other federal instruments (research funding programmes, Digital Pact, NUM initiatives) and is intended to be coherent with EU regulatory frameworks and initiatives (including the EU AI Act and other European research and infrastructure programmes). The plan references existing BMBF and inter‑ministerial measures and seeks to align funding, infrastructure and competence-building actions with broader national and EU strategies to avoid duplication and to maximise leverage of European resources.

International Alignment

International alignment is central: the Action Plan foregrounds close European cooperation to achieve "AI Made in Europe" and technological sovereignty. The BMBF commits to convening a high-level European AI workshop and to align national investments with European research infrastructures (EuroHPC / European HPC initiatives), the European AI On-Demand platform and other EU-level instruments. The plan stresses that regulatory choices should be compatible with EU initiatives (and refers implementationally to the EU AI Act and Commission guidance) while promoting European models for trustworthy, competitive AI.

Implementation Timeline

Sources and References