EA, AI and strategy

Something I was thinking while riding subway to the office today. Maybe something worth discussing together. 

For a long time, it has been wise to design enterprise architecture (EA) (and also enterprise systems) in such a way that they can accommodate (or adapt to) changes in corporate strategy. The reason for this is simple: it is way easier to change strategy than it is to change enterprise architecture and system. Thus, the main difficulty in EA lies in understanding what are the limits of possible strategies, what kind of new business, products and service the corporation may choose to want to have. 

This has required good relationships with business leaders, discussions with the CEO and maybe even the board (or its individual members). It has also required good knowledge of both EA and also business, and also theoretical knowledge of business strategies. It has never been enough to only know technical stuff, or to be able to model the current or future states. And, especially, the TOGAF way of deducing EA from corporate strategies has not been possible. EA and strategy are intertwined. 

Now, AI (LLMs, machine learning) is becoming more important. Does it change the relationship between strategy and EA? Do we need to take AI into account in addition to strategy? It seems that AI is developing and maybe maturing much faster than any corporate strategy can. Does this mean that in EA we must consider AI, to make sure AI can be adopted quickly? 

How can we make EA more “AI proof” while also making sure that we still do not exclude any plausible changes in corporate strategy? Who should we discuss with in the corporation - I assume the CEO is not best expert.

Has anyone been thinking about this?