A small portion of the German workforce, approximately 20%, possess the skills to effectively utilize artificial intelligence.
In a recent survey conducted by Bitkom, it was revealed that many working people in Germany anticipate significant changes in their workplace due to Artificial Intelligence (AI). The survey, which polled 1,005 individuals aged 16 and over, did not specify the number of professionals who have AI training available but have not taken it up.
One in three respondents (33 percent) believes AI could replace their managers, while only one in five professionals in Germany have received AI training at work. Moreover, a further 6 percent of professionals have AI training available but have not taken it up. Most German employees have not been offered training in handling artificial intelligence by their employers.
The survey did not mention whether companies that do not offer AI training may be violating current laws. However, the EU AI Regulation (AI Act), which came into effect in February 2025, requires all companies using AI to ensure that those involved have sufficient AI competence.
Under the EU AI Regulation, there is a clear requirement for AI literacy training for employees and all personnel interacting with AI systems, including contractors and service providers. The key points of the AI competence training requirements include:
1. Scope: The AI literacy obligation applies to all users of AI systems within an organisation, not just employees. 2. Content of Training: Training must go beyond basic operational knowledge. It should cover general understanding of AI capabilities and limitations, the ability to critically interpret AI outputs, assessment of legal and ethical implications, awareness of risks such as hallucinations in generative AI applications, and more. 3. Tailoring: Training programs must be tailored to the organisation’s role (whether AI provider or deployer), the risk profile of the AI system in use, and the technical competencies of staff. 4. Integration: AI literacy initiatives should be integrated into broader risk management and compliance frameworks, with bespoke policies and procedures complementing training rather than relying solely on user manuals or passive documentation. 5. Proportionality: The extent and robustness of training depends on the risk level of the AI systems deployed — higher-risk systems require more comprehensive literacy programs. 6. Timing and Enforcement: These training and literacy requirements have been in effect and are enforceable since February 2, 2025. The European Commission is expected to provide further guidance, including a Code of Practice for general-purpose AI, expected by August 2025. 7. No Prescribed Curriculum: The AI Act does not specify exact training curricula or methods, leaving flexibility for organisations to design programs that suit their context and obligations.
Ralf Wintergerst, Bitkom President, emphasised the importance of knowing how to use AI tools correctly and understanding the technology's possibilities, limitations, data protection, and data security. He stated that AI makes work tasks easier and more efficient.
Companies that do not offer AI training may be violating current laws, and those that do not currently use AI are encouraged to consider offering relevant training, as employees may use private AI apps for work purposes. The EU AI Regulation applies not only to a company's own employees but also to freelancers, temporary workers, or service providers.
- Considering the EU AI Regulation, companies in Germany that have not offered AI literacy training to their employees and all personnel interacting with AI systems might be breaching current laws.
- With the EU AI Regulation in place, it is crucial for businesses to provide training that goes beyond basic AI operations, covering aspects such as understanding AI capabilities and limitations, interpreting AI outputs, assessing legal and ethical implications, and implementing such training within broader risk management and compliance frameworks.
