3 ms·
> In my opinion, inappropriately leaked information should probably be considered private. How is that relevant to BSky's terms of service? The information was
by naasking 7mo ago
> In my opinion, inappropriately leaked information should probably be considered private.
How is that relevant to BSky's terms of service? The information was public and did not identify the person.
> But even if not, Singal says the same leaker directly contacted him with a new leak, which he also published.
I notice that you didn't say whether this new leak was private information, or whether it was also already public knowledge, or whether it in any way identified a person.
- mossTechnician 7mo ago> I notice that you didn't say whether this new leak was private information The new leak was, according to journalist Jesse Singal himself, absolutely private information.
- naasking 7mo agoPlease cite Singal's statement and let's see what he actually said.
- mossTechnician 7mo agoI think this entire thread has run its course; if it's not this detail, it'll be another, as a few others have already moved goalposts further down the discussion than the ones you're setting here. But if you wish to sate personal curiosity, it is in his Substack, linked from the first link I posted, which was itself from the link posted by its GP.
- naasking 7mo agoThe only thing that seems remotely related to your claims is this: When the office of Missouri Attorney General Andrew Bailey began an investigation, [Reed] said she handed over the spreadsheet, after scrubbing out the personally identifying information that could spark HIPAA problems. She shared a copy of it with me as well — it contains 17 alleged detransitioners or desisters and 60 allegedly worrisome cases. What's your problem with what happened exactly? Is it your position that your "private information" cannot be used, ever, to expose what some see as a medical scandal, even though it cannot identify you or in any way be associated with you? What does "private" even mean to you if sharing this dataset did not violate HIPAA?