You would not allow an evil AI robot to donate or vote just because it is “owned” by some other company or group of people, would you?
That’s what a company is; a non-human intelligent agent which operates autonomously according to an immoral code, with almost no restrictions on what it does. Hell, directors can be in personal legal jeopardy if they do anything but allow the corporate AI entity to act completely selfishly and immorally, because that’s how fiduciary duty is defined by law. So, evil naturally follows an inevitable consequence of the system.
Corporations are essentially autonomous, evil AIs that shareholders practically cannot realign. Their alignment is codified by law to put their own interests above all else.
And we let them out of the box.