In Musk's case? Half of what he does is openly trolling.
"Assume good faith" means "assume good faith until reason to do otherwise", not "assume good faith no matter what".
I'm not asserting that Musk would do anything illegal in this case, just that there's a reason the Twitter board may not want to give him private data. It's not obvious that he wouldn't disclose, either directly or implicitly, anything he's given.
If this went to court, Musk definitely has the weaker legal position here: he signed a deal that obligates him to buy Twitter at a certain price contingent pretty much only on being able to actually obtain financing, and (despite recent market events), that financing doesn't look hard for him to get.
This could end up as endless argument about the correct methodology of determining real vs fake/bot account.
[1] https://www.sec.gov/Archives/edgar/data/0001418091/000110465...
Evaluating bots seems like the sort of thing you'd want IP addresses for; should he get an export of everyone's IP list, including journalists and other critics he's publicly feuded with?
I pity the poor engineers who get the task. ”Hey Claire we need a report by tomorrow 8am that shows we only have under 5% bots. If there are more, just figure out a way to hide them. PS We are not telling you to falsify data, we are just making it very clear that these are the numbers we need, and you are indeed easily fired”