Of course, many of these posts are public and thus were already accessible to any FBI agent willing to do some detailed police work. The difference is that now the agency can search and sort all of Twitter’s posts to find trends — or zero in on particular individuals, including those who have set their feeds to private.
The FBI got this access through a contract with Dataminr, a company partly owned by Twitter itself and the one company able to give users access to the full “fire hose” of Twitter data. (The publicly-available API only allows access to about 1 percent of all tweets.)
The bureau says it needs real-time access to all of Twitter to catch terrorists and other criminals. But needless to say, the overwhelmingly vast majority of Twitter users are not criminals, and their tweets will be subject to search anyway.
Moreover, in a grim irony, getting the full fire hose of Twitter data may actually make the FBI’s investigations less accurate.
You see, in theory, getting every possible piece of information sounds like a good investigatory method, but in practice, this much data crushes law enforcement officers, bogging them down in details that more traditional, limited surveillance would skillfully avoid (emphasis added):
By flooding the system with false positives, big-data approaches to counterterrorism might actually make it harder to identify real terrorists before they act. […] “The bigger haystack makes it harder to find the needle,” [Republican Iowa Rep. James] Sensenbrenner told me.
Thomas Drake, a former N.S.A. executive and whistle-blower who has become one of the agency’s most vocal critics, told me, “If you target everything, there’s no target.” Drake favors what he calls “a traditional law-enforcement” approach to terrorism, gathering more intelligence on a smaller set of targets. Decisions about which targets matter, he said, should be driven by human expertise, not by a database.
The FBI’s plans for Dataminr use also seem to run afoul of Twitter’s own terms of service, as Russell Brandom at The Verge explains:
[T]he FBI contract seems to violate a key clause in Twitter’s Developer Agreement, which specifically forbids using the provided data to “investigate, track or surveil Twitter’s users.” In practice, that has often meant banning third-party companies found to be reselling data. […]
This isn’t the first time Dataminr has run up against Twitter’s anti-surveillance clause. In May, Twitter revoked CIA access to Dataminr, a move that was taken as part of a larger ban on US intelligence agencies using the product. “Data is largely public,” Twitter said in a statement at the time, “and the U.S. government may review public accounts on its own, like any user could.”
So it’s possible the FBI could lose this huge new spying capability, but personally I don’t think it’s likely.
Why not? Well, it turns out Dataminr already has contracts with the TSA and the Pentagon. It’s difficult to imagine those agencies — especially the ever-dastardly TSA — aren’t using their Dataminr access to “investigate, track or surveil Twitter’s users.”