The FBI Wants To Be Your Facebook Friend

(Mis)Uses of Technology

from the please-share-this-post,-thx,-your-friendly-g-men dept

Undoubtedly, the FBI has always surveilled the open web, looking for persons or phrases of interest. It’s just going to get a whole lot better at doing it. And it’s going to spend millions of your tax dollars to make it easier to place your public internet interactions under its social media-focused microscope. Aaron Schaffer has the details for the Washington Post.

The FBI has contracted for 5,000 licenses to use Babel X, a software made by Babel Street that lets users search social media sites within a geographic area and use other parameters.

The contract began March 30 and is worth as much as $27 million. The FBI has already agreed to pay an IT vendor around $5 million for the first year of the contract, procurement records indicate. The contract has not previously been reported.

Babel X is one of Babel Street’s products. Babel Street has made previous headlines here at Techdirt for selling location data gleaned from cell phone apps to federal agencies (including CBP, ICE, US Secret Service and the US military), allowing them to bypass warrant requirements erected by the Supreme Court’s Carpenter decision.

Babel X targets the internet, focusing on publicly viewable posts hosted by a large number of social media services. The licenses obtained by the FBI target Twitter, Facebook, Instagram, YouTube, LinkedIn (?), VKontakte, Telegram, and “Deep/Dark Web.”

The FBI’s wishlist runs deeper than the online services listed above. It also would like access to posts on 8Kun, Discord, Gab, Parler, Reddit, Snapchat, TikTok, and Weibo, but doesn’t consider lack of immediate access to be a dealbreaker.

The FBI will be performing lots and lots of searches, some possibly constrained by geofences tossed up around areas of interest.

In contracting documents, the FBI estimates that its 5,000 licensees will run around 20,000 keyword searches every month, though it cautioned that that’s “merely an estimate.”

The FBI only wants access to content it can obtain without logins or court orders. That’s still going to be millions of posts depending on the keywords used or the area searched. On top of keyword searches, the FBI wants to be able to capture and analyze “emotions and sentiments,” which will apparently necessitate the use of “emoji searches.”

This will put the FBI on the receiving end of an extremely productive fire hose. Hence the need to spend millions not just on content acquisition, but the tools to make sense of the shit tonnes of data the FBI will obtain with Babel Street’s tools.

Now, it’s been long understood that anything viewable by members of the public can be viewed by law enforcement. But some courts have had problems with the government assuming powerful tools, that exponentially increase surveillance powers by becoming the equivalent of thousands or millions of unblinking eyes, are free of constitutional concerns. The courts have concerns even if law enforcement often does not.

On top of that, there’s the “reasonable expectation of privacy.” While people may understand that others may see their posts if they maintain a public account, they don’t reasonably expect government employees should be able to eavesdrop at will, much less collect their posts in bulk and subject them to powerful analytic software.

This will lead to self-censorship and the mistaken targeting of people who routinely deal with issues that align with government interests, like activists, journalists, rights groups, FOIA enthusiasts, and critics of government activity.

So, some of these people will either take their accounts private or steer clear of discussing terms the FBI might find interesting, just to steer clear of future hassle. That’s not free speech, as the ACLU’s Matt Cagle points out:

“The First Amendment protects online speech, period,” he said. “People should not have to exercise their free speech behind privacy settings in order to avoid being surveilled.”

It appears the FBI is going to collect it all. The documents say it wants to “analyze past events.” But it also wants to run “persistent” searches every eight minutes and feed these results into its Babel Street enabled predictive policing program.

While the FBI has offered a statement stating basically that this is legal, useful, and probably will do cool anti-terrorism stuff, it did not offer any assurance that policies were being crafted to regulate collection and retention of this content. Nor did it say anything about the federally required Privacy Impact Assessment it needs to hand in before deploying these Babel Street licenses. It will probably take the intervention of congressional oversight to get better answers than the ones the FBI is willing to provide voluntarily. And while we wait, the FBI will be subjecting free speech to government scrutiny.

Filed Under: babel x, fbi, social media, surveillance
Companies: babel street