[_delirium]

The most famous case is probably Neal Horsley, who was found liable (though only civilly) for publishing the home addresses of abortion doctors, some of whom were later harmed. He was sued for contributing to the attacks and lost. These were correct addresses and it was true that they were abortion doctors, but a court nonetheless considered their publication to constitute an implied threat of harm to the doctors in question. There are some less dramatic cases involving the "public disclosure of private facts" tort, against which truth is not a defense (though an address would not usually be considered a private fact).

In the traditional (pre-automation) era intent played a large role. This worked out okay in most cases, because harmful things (like Horsley's site) tended to only show up as a result of malevolent intent. Sites with lists of abortion doctors' home addresses didn't just show up innocently, so if nobody had malevolent intent, they wouldn't show up. Now if algorithms are just throwing up lots of things, there is probably not malevolent intent on the part of the algorithms, but the same harms can result. So different legal systems are trying to figure out what to do about it.

[Crito]

Getting nailed in civil court is a completely different animal than criminal charges. You can even get nailed in civil courts for things that criminal courts found you not guilty of (see: OJ Simpson).

[_delirium]

But that's what the discussion here is about as well. Google's "right to be forgotten" loss was also a civil case, not a criminal one.