It seems to be accepted wisdom in the business world that reference letters for former employees should be extremely terse. They should confirm that the employee worked there, and essentially nothing else:
To whom it may concern: Ellen Ripley was employed by the Weyland-Yutani Corporation from June 2137 to September 2139. Signed, C. Burke, manager."
The standard reasoning is that if the letter contains something unfavorable and the employee is turned down for a future job, they might sue their former company for libel, claiming the unfavorable statement was a lie which damaged their career. In order to avoid the possibility of such a legal battle (which, the reasoning goes, could be very expensive, even if the company wins), the company tells its managers to write letters with no content, so that there's no chance of them containing something actionable.
This question on Workplace Stack Exchange, Is there any evidence that giving references for former employees is inherently risky?, attests to this practice, and the answers provide some suggestion that the company's fears are justified.
On the other hand, in academia, detailed and informative recommendation letters are the norm. They often run to multiple pages, and contain specific information about the candidate's history and activities at the institution, as well as the writer's (supposedly honest) subjective assessment of the candidate's strengths, weaknesses, and potential. This is not only common but effectively mandatory; a minimal recommendation letter of the kind described above would immediately consign the candidate's application to the nearest wastebasket.
This would seem to be just the sort of thing that fills corporate counsel with horror, yet we do it every day. No academic employer of mine has ever told me not to do so; for that matter, I can't say that I've ever received any official guidance, one way or the other, on writing recommendations. Nobody in academia seems to be concerned that writing a letter that could be construed as less than favorable could result in legal consequences. So how are we getting away with it?
Are universities treated differently under the law, making them less vulnerable to such threats? Or do they willingly accept the legal risks in order to make the world a better place by providing actual information about their former students/employees? Or is everyone ignorant of the risk of trouble? Or is there something I have not thought of?
To forestall a couple of objections: I know that academics usually try to write only positive letters, and decline to write if they have nothing good to say, but that's evidently not enough in the corporate world. Anyway, candidates often end up damned by faint praise. Also, I know that we send letters in confidence, with the understanding that they won't be shown to the candidate, and sometimes this is backed by having the candidate sign a waiver, but I have to believe that a sufficiently determined and litigious candidate could get access anyway.
(I know this question may sound rhetorical, but I ask it in seriousness, and hope to learn something from the answers. My context is the US, but if things are substantially different in other places, that would be interesting to know as well.)