If an individual acted in a way consistent with the "rogue AI agents" that some of your name-checked companies claim their AIs acted, then that individual would be pursued by law enforcement. (For example, Aaron Swartz.) That these companies are allowed to misbehave without repercussion is a form of letting the gains belong to a few, while the risks are socialized among the many. Another example would be "Too Big to Fail" bailouts after 2008. Those bailed out companies should belong to the American populace now, paying us dividends, so that the employees could keep their livelihoods. Alternatively, but I think this is a more callous approach because of knock on effects, they could have been allowed to fail. That they were not is clear and convincing evidence to most people of what I think the prior commenter is referring to as "corporate socialism."