In line with the first major milestone of the EASA Artificial Intelligence (AI) Roadmap 1.0, EASA says this concept paper presents a first set of objectives for Level 1 Artificial Intelligence (‘assistance to human’), in order to anticipate future EASA guidance and requirements for safety-related machine learning (ML) applications.
The goal of this document is twofold:
- to allow applicants to have an early visibility on the possible expectations of EASA with respect to the implementation of AI/ML solutions.
- to establish a baseline for Level 1 AI applications that will be further refined for Level 2 and Level 3 AI applications.
This document has been matured over the last 1.5 years and underwent several stages of consultation including a 10 weeks period of public consultation from April to June 2021. It covers only an initial set of AI/ML techniques and will be enriched with other advanced techniques, as the EASA AI Roadmap is implemented.
In 2022, this document will be further developed to include the provisions for more advanced types of AI, the so-called level 2 AI applications, involving a higher degree of collaboration and even teaming between the human end user and the AI-based systems.