This is the issue with AI: it is corporatized, and it is weaponized for capitalism.
We already are at the boundary of insidious total immersion advertising for psychological manipulation from the last five decades of mass media since the mass adoption of television.
But AI is simply another level, and it isn't going to be "early google don't be evil". That was the outgrowth of the early internet. From protocols that were build to be sensible, not commercial weaponized protocols.
AI, human-computer-neural interfaces, and other types of emerging deep-intellectual-penetration products are all FULLY WEAPONIZED for commercial exploitation, security dangers, propagandization, and zero consumer privacy. They are all being developed in the age of the smartphone with it's assumed "you have no privacy, we listen to everything, track everything, and that's our right".
It's already appalling from the smartphone front, but AI + VR + neural interfaces are just another level of philosophical quandry, where an individual's senses, the link to "reality", is controlled by corporations. Your only link to reality is the vague societal and governmental control mechanism known as "money".
The internet protocols (the core ones) were built for mass adoption by the world with a vision for information exchange. They were truly open. They weren't undermined by trojan horses, or an incumbent with a massive head start that is dictating the protocol to match their existing products.
AI+VR is the same new leap information transmission, but it is NOT founded on good protocol design. By protocols I mean "the basic rules". There are no rules, there is no morality, and there is no regulation. Just profit motives.