This is largely due to the age old fact that corporations rarely make decisions based on actual data, introspection, and good judgment. Usually the decision is made first and then the justifications are invented afterwards.

In this case, every executive is terrified of being "left out" in the AI race. As we saw with the mass layoffs across companies, most of CEO decision making is just adhering to herd behavior. So it is literally better for execs to have shoveled a shit ton of money into 'strategic' AI initiatives and have them fail than potentially deal with the potentially remote chance of some other exec or company succeeding with 'AI enabled transformation'.

What makes it even more fun is that nobody really has a good understanding of how to measure the ROI of AI. Hence we have people burning a lot of money due to FOMO and no easy way of measuring the outcome, which is usually how the foundations for good Ponzi schemes are laid.

This is unlikely to end well. However, as usual, it's us, the common plebs, who will suffer regardless of outcome.