Digital transformation in combination with service ecosystems exploit and incorporate innovative technologies, such as artificial intelligence, Internet of things or data analytics. For most enterprises, new digital business models and participation in service eco-systems lead to severe changes of the enterprise architecture (EA) and the need for methodical support to systematically perform the resulting change process. The focus of this paper is on the implementation of AI applications in organizations. Based on the analysis of industrial case studies, our observation is that different kinds of AI applications require different prerequisites in an organizational IT landscape, some of which can be found in an EA model, and some enterprises intend to use AI but are not prepared for it. The work investigates, what information can be harvested from EA models to support requirements engineering and the evaluation of organizational readiness for AI planning and implementation. The main contributions of our work are (a) an enhanced and updated literature analysis on EA use for AI introduction, (b) an analysis of differences in requirements of different kinds of AI applications, and (c) an improved AI context analysis method prepared for addressing these differences. The improved method exploits insights gained in our work on what information regarding the requirements can be extracted from EA models.