You're right, and it goes much deeper. These corporations effectively own enough of our data to trivially deanonymize and target people. An organization of people who believe they are morally justified in they cause are liable to abuse such information, regardless of which side they belong to.

That's the most terrifying part to me as a dev. We don't even need neural networks to search petabytes of social media content for keywords. If we allow corporations to become politicized then we open ourselves up to unprecedented abuses.

You're right that Internet-ish companies are a special risk. But even run-of-the-mill organizations seem to be drinking the Kool-Aid these days. I sat today in an all-hands anti-racism videoconference. It was bland enough, though it would at one time have made a good SNL skit. (It was a bunch of woke white women declaiming our failures.) Committees were proposed and also the hiring of (even) more D&I consultants and personnel. As usual, no actually useful action was proposed, but we have once again checked the box.