tekom Europe strongly supports the EU’s proposed AI Act but stresses the need for clarifications and improvements regarding technical documentation and instructions for AI use. The association advocates for standards that ensure transparency, ethics, and user safety.
The AI Act aims to regulate the transformative impact of AI on industries and society, ensuring safe and just integration. However, details regarding the impact of AI, particularly GenAI, on technical documentation development remain insufficiently addressed. The AI Act addresses technical communication, including technical documentation and instructions for use, tailored to different AI classifications such as prohibited, high-risk, and low-risk categories.
tekom Europe’s position paper poses several concerns and requests regarding technical documentation within the AI Act.
Here is a summarized list of tekom Europe’s concerns and requests regarding the AI Act:
- Lack of specificity regarding instructions for use for General Purpose AI (GPAI), such as GenAI
- Need for more detailed information on the production of technical documentation, including:
- Creation of simplified technical documentation forms
- Role of AI in generating technical documentation
- Lack of consideration for in-house AI systems developed by organizations for their own use
- Uncertainty regarding the completeness of instructions for use for foundational AI models like GenAI, particularly regarding:
- Inclusion of engineered prompts
- Levels of accuracy required
- Questions about the autonomy of AI in producing technical documentation and the necessity of human oversight
- Uncertainty about the necessity of conformity assessment and EU-certified technical documentation for isolated or in-house AI systems, suggesting a need for specific guidance tailored to these developments
Read the full position paper here:
POSITION PAPER OF THE TEKOM EUROPE ON THE EUROPEAN ARTIFICIAL INTELLIGENCE (AI) ACT