People don't buy merchs of these well-established IP for their astonishing production value. And especially not for how cheap they are.
* deprecating people's models so that they have to buy new ones
* making any number of rules changes that were widely hated
* making lore changes that were widely hated
They aren't going to lose customers because some other company is using AI. They effectively don't have any competition, because people love the Warhammer settings and want to play games set in them.
That is a false dichotomy. Eschewing AI may actually provide a competitive advantage in some markets. In other words, the third choice is to pivot and differentiate.
Their worlds are their monopolies. Worlds that now have multi-decades worth of lore investment (almost 50 years now I think).
Just because someone else can make cheaper little plastic models doesn't affect GW in the slightest. Or pump out AI slop stories.
The Horus Heresy book series is like 64 books now. And that's just a spin-off. It's set way before when 40k actually is set (10,000 years).
With so much lore they need complicated archiving and tracking to keep on top of it all (I happen to know their chief archivist).
You can't replace that. I only say all this just to try and explain how off the mark you are on understanding what the actual value of the company is.
I live in Nottingham where GW is based, another of my friends happens to have a company on an industrial estate where there are like 3 other tabletop gaming companies. All ex-gw staff.
You could probably fit all their buildings in the pub that GW has on its colossal factory site.
You used to know people who worked at Boots, which used to be the big Nottingham employer. Now days, I know more people who work at GW.
Plenty of people use proxies, too. There's places that do monthly packs of new STLs that could be an entire faction army, and there's long been places that sold "definitely not Space Marines and Sisters of Battle" minis too.
They don't have a threat of anyone overtaking them at current, but AI making alternatives in this vein even cheaper could eat away at portions of their bottom line.
As an aside, I am somewhat disappointed that Battletech's appeal to the mainstream is largely down to the Mechwarrior games which have minimal lore.
There is so much more that could be done. But the current owners seem to be pretty poor at translating all their paperwork stories for the modern crowd.
I mean, I really doubt that making tabletop games with AI slop would grant anyone a substantial competitive advantage.
They will definitely start using AI when their competitors do to the point that they gain a substantial competitive advantage. Then, at least in a free market, their only choices are to use AI or cease to exist. At that point, it is more survival bias (companies that used AI survived) rather than profit motive (companies used AI to make more money).