There's also some confusion at the start because DOJ asked for other possible accounts that Trump might have used as a sockpuppet/backchannel means of communication and Twitter seems to either not be able to discover that, or can't find it.
They clearly CAN, but don't want to. Their (X's lawyers) objections in the transcript are ridiculous (they claimed they don't have an "affiliated account" category - like really?!?) and when pressed for timeframes suddenly they can produce "impossible" data in less than 24h.
If @realdonaldtrump and @johnbarron are both tweeting from the same IP in the same time frame, could Twitter make the connection that both are the same user? Was there enough backtracking stored in their systems to know that?
Not saying it's a viable technique but I get the idea that judges hate warrants that vague.
Legally, it seems asking for a company to run essentially joins across their DB tables is accepted practice. Otherwise, how could warrant be obtained from google for all devices within a specific geolocation? Requires at least a few DB queries I'd imagine.
They are pretending that a database search for an IP address, for example, in two different tables is a big challenge and requires them to "make" the data because they don't have all the matching IP addresses pre-queried.
Twitter is likely doing this because musk is now in charge, and wants to help the criminal conspiracy by hiding and denying the legally requested evidence
https://www.nytimes.com/2023/08/15/us/politics/trump-twitter...
I'm pretty sure this is record of what it took for the special counsel investigation to get access to the DMs of @realDonaldTrump.