The root problem here is that technology is being sequestered and locked down to large centralized entities that can be blamed, and it doesn't help that they're

The root problem here is that technology is being sequestered and locked down to large centralized entities that can be blamed, and it doesn't help that they're intrinsically risk adverse. Information provided by web services is the company doing something, and said companies obviously don't want to be found responsible for illegal acts. So, can't say anything that might influence elections, can't say anything that might help or induce someone to kill themselves, can't say anything that might help someone avoid suffering the medical cartel or attorneys' guild, and so on. Whereas if this were software being sold/provided at an arms length, the creator would no longer be a party to its use and there would be very few theories of liability to hold its creator liable for illegal uses (without the creator explicitly warranting or encouraging it for those uses).