National AI Strategy (Germany)

3.1 Strengthen research to drive innovation

Germany will strengthen AI research through a network of at least twelve competence centres and application hubs with competitive working conditions, covering general methodology, sector-specific applications, and interdisciplinary clusters, while establishing a national research consortium, creating at least 100 new AI professorships, supporting early-career researchers and German experts' return, and establishing a German-French research network with bilateral funding. The strategy prioritizes technological sovereignty in hardware and software, research into explainability and privacy-preserving technologies including differential privacy, ethical AI design integrated into standardization, and application development in biotechnology, autonomous driving, agriculture, sustainability, health, civil security, biomedical data analysis, care robotics, space technologies and earth observation, with federal agencies coordinating transfer structures and integrating regulatory expertise early in research and development for high-regulation domains. The government will build a teaching and learning platform on AI, support data sovereignty through data integration centres at university hospitals, establish two new competence centres for autonomous systems in human-hostile environments, and review funding procedures to improve accessibility and simplify administration.

3.2 Innovation competitions and European innovation clusters

The government will establish AI as a priority in a planned agency for breakthrough innovations and will create a European innovation cluster to fund collaborative research projects over five years with other European partners in science and industry. The government also considers establishing an Important Project of Common European Interest in AI with other European partners if needed.

3.3 Transfer to business, strengthen mid-market

The federal government aims to strengthen German and European business competitiveness through widespread adoption of artificial intelligence, with targeted support for mid-sized enterprises, by launching a transfer initiative to accelerate conversion of research into market-ready products, supporting patent applications, expanding science-business networking, deploying at least one thousand AI trainer contacts annually through Competence Centre Mittelstand 4.0, and supporting establishment of test fields and real-world laboratories for technology trials. The federal government will promote business cooperation within competition law, support consortiums for global competitiveness, establish a Competition Law 4.0 Commission to develop policy recommendations by autumn 2019, and conduct regular AI penetration monitoring across sectors to measure internal company application, sales of AI applications, and the origin and use of data and hardware.

3.4 Awakening and supporting startup momentum

Germany commits to increasing startup formation around AI-based business models by improving access to venture capital during growth phases through measures including doubling EXIST programme funding in 2019, expanding venture capital and venture debt offerings via a Tech Growth Fund Initiative, establishing KfW Capital to increase annual venture capital and venture debt investment to 200 million euros by 2020, launching an online startup platform with 400 partner advisors, and strengthening AI competence through the Digital-Hub-Initiative. The Digital-Hub-Initiative will intensify exchange on AI-relevant topics tailored to startup and established company needs in individual hubs. Germany established a FinTech Council to address AI-driven changes in finance including more efficient core banking processes and new data-driven business models, and the financial regulator published a 2018 study analysing opportunities and challenges of big data and AI for financial services supervision and regulation.

3.5 Shaping work and labor market change

Germany adopts a human-centred approach to AI's impact on employment through a National Continuing Education Strategy, the Qualification Opportunities Act providing retraining and wage subsidies (substantially higher for small and medium-sized enterprises than large enterprises), skilled worker strategy with domestic, European and international potential, a German AI Observatory monitoring AI effects on work, and European and transatlantic dialogue on intelligent systems deployment. The government will clarify that works councils' existing co-determination rights under the Works Constitution Act cover AI deployment, amend Section 90 of that Act to explicitly include planning and use of AI applications, clarify that Section 95 co-determination rights apply to AI in recruitment, examine a separate employee data protection law for workers' fundamental rights and data self-determination, establish works councils' initiative rights on retraining, and examine auditing of AI in workplace contexts on employment, technology design, human-machine interfaces, health and data protection. Support includes funding for workplace experimental spaces testing AI applications under the New Quality of Work Initiative, regional innovation centres in each eastern German state offering training to employees, works councils, managers, trainers and self-employed workers, and a House of Self-Employment supporting solo self-employed and platform workers initially in eastern states and then nationwide.

3.6 Strengthening education and acquiring expertise

Education and professional training must adapt to digital transformation and AI by promoting independent thinking and examining ethical, social and technical dimensions alongside technology-focused instruction, with a comprehensive approach across schools, vocational training, universities and workplace learning. The federal government will promote digital infrastructure in all schools and vocational schools through DigitalPakt Schule during this legislative period (conditional on amending Article 104c of the Basic Law), ensure all students enrolled from school year 2018/2019 onwards acquire comprehensive digital competencies by the end of schooling, support programs like Lernfabriken 4.0 for hands-on experience with automation technology, and integrate AI knowledge into vocational training including in specific sectors like healthcare and food supply. The federal government will fund educational research through the Digitalisierung II guideline to develop concepts for educational processes utilizing digital media, investigate AI's role in learning processes, explore options and limits of learning progress diagnostics for self-directed learning, and examine applications of Learning Analytics while considering ethical and data protection concerns.

3.7 Using AI for public functions and administrative capacity

The federal government aims to deploy AI in public administration to improve efficiency, quality and security of administrative services, expand publication of administrative data under open by default principles, and develop measures for threat detection and prevention in internal and external security with adequate control and transparency. The government will assess impacts of potential AI applications it may exclude for political, legal or ethical reasons and identify suitable areas for security authorities to deploy AI for pattern recognition in large data analysis, criminal prosecution, threat prevention, child protection online, combating abuse material, and social media forensics for profiling, with decisions remaining with human officials and subject to personality rights and specific legal conditions.

3.8 Making data available and facilitating use

The government aims to substantially increase the quantity and quality of usable data for AI while protecting constitutional rights including privacy and informational self-determination, and will examine establishing a national trustworthy publicly accessible data and analytics infrastructure as a joint venture. The government supports European data space initiatives, will develop antitrust guidelines for data partnerships, make publicly funded research datasets available to third parties, expand access to earth observation data, promote privacy-compliant training datasets and research into anonymisation and synthetic data, encourage public administration data as open data, harmonise mobility data access, and establish guidelines for data protection compliant AI development in the automotive sector including methods for anonymization and pseudonymization. In the health sector, the government will enable data from care and billing processes to be used for machine learning while respecting data sovereignty and patient rights, establish formats for data exchange between care and research, and support projects bundling distributed data across research and care provision.

3.9 Adjust regulatory framework

The government aims to ensure that AI applications respect Germany's democratic constitutional order and protect fundamental rights through ethics by, in and for design approaches, and will review European and national legal frameworks for gaps regarding personal and non-personal data use, algorithmic discrimination, security standards, and transparency of AI systems. Those affected by automated decisions must be able to access decision grounds including criteria, objectives, logic and personal data used in clear, accessible language; transparency, traceability, freedom from discrimination and verifiability must be guaranteed throughout AI development, programming, introduction and use, particularly for systems making or drawing conclusions for decisions and also where no personal data are involved. The government will review establishing or expanding public agencies and private audit institutions to monitor algorithmic decisions, develop auditing and impact assessment standards, and require full disclosure of AI and algorithmic decision-making processes to oversight systems without requiring disclosure of trade secrets.

3.10 Setting standards

The government is responsible for setting the regulatory framework and promoting standards and norm-setting at national, European and international levels through DIN/DKE organizations, with key priorities including standardizing AI terms and classifications (autonomy dimensions, learning autonomy, associated risks), ethical standards (ethics by design), and reviewing existing standards for AI compatibility. The government will consider supporting expert participation, particularly from small and medium-sized enterprises and startups, in international standardization procedures and will develop a roadmap with DIN on AI-related norms and standards. The government will coordinate with EU-wide efforts to strengthen European interests in international standardization bodies, including input from the data ethics commission on ethical matters, to better enforce European values internationally without improper bloc voting.

3.11 National and international networking

The government will coordinate AI strategy measures with other government strategies including digitalization, sustainability, industry 4.0, health sector digitalization and climate protection, assess feedback effects across policy areas, and cooperate with the Bundestag's AI inquiry commission, while strengthening EU cooperation on digital single market conditions and AI strategy implementation through digital hubs, excellence centres, coordinated research programmes and expert exchange by 2027, and pursuing international collaboration through G7, G20, OECD and United Nations forums to develop joint research with nations such as the USA, Canada, Israel and certain Asian countries, establish AI observatories, develop shared guidelines and standards, and build AI capacity in developing countries. Germany will support developing and emerging countries in building scientific and technical expertise through education and open-source and open-data approaches to prevent discrimination and ensure suitability, support adapted policy frameworks, and promote digital and media competence and critical reflection to enable meaningful engagement with AI. Germany must increase investments in technology impact assessment, advance research on data portability, interoperability and consumer-enabling technologies, ensure science and civil society can independently contribute to societal discourse without dependence on AI company funding, and lead public discussion on which human competencies and independence should be preserved as AI develops.

3.12 Societal dialogue and framework development

Germany will conduct intensive societal dialogue and establish participatory opportunities involving the federal government, educational institutions, cultural institutions, centres of expertise, states and municipalities to anchor AI culturally, ethically, legally and institutionally, ensure public understanding and acceptance, address information gaps through broad media and technology education drawing on civil society experience, ensure participation of all population groups including those with low digital engagement, and conduct interdisciplinary and transdisciplinary dialogue on AI's socioeconomic and sociocultural effects and technology impact assessment. The government will establish a Future Fund for Digital Work and Society to promote broad education and multidisciplinary social technology design, initiate interdisciplinary science dialogue on socially responsible technology development and human-centred AI innovation, support dialogues between science, business and civil society, identify beacon projects, run idea competitions, develop funding programmes for dialogue-oriented education projects and experimental spaces in education, culture, administration and civil society, support municipalities in establishing citizen labs, promote education and awareness campaigns, and ensure conditions that maintain diversity and guarantee space for cultural and media freedoms. The government will develop the Learning Systems Platform into an AI platform bringing together experts from science, business, politics and civil society to discuss opportunities, challenges and conditions for responsible AI deployment through specialized working groups and derive scenarios, recommendations and roadmaps including ethical and legal clarifications, accompany social partnership dialogues on sustainable AI integration in the workplace through expert forums in the Federal Labour Ministry and the New Quality of Work Initiative including experimental spaces and knowledge transfer, and establish a communication strategy for AI emphasizing human-centred technology paths, public benefit orientation, better visibility of existing support services, understandable AI representation to business and society, and enhancement of Germany's international AI image.