The FBI is investing millions in social media surveillance software

Back view of young man with headset watching surveillance footage and map on computers and speaking ...


The FBI's total investment in Babel X.

Washington Post


Jokes about the FBI watching your every move through webcams are about to get a little more real. The Bureau just signed a $27 million contract to use Babel X, AI-powered software that trawls social media for specific content. It’s exactly the kind of surveillance software privacy activists are constantly warning the public about.

The contract, which includes 5,000 licenses for Babel X, began on March 30, The Washington Post reports. About $5 million of the FBI’s contract will be paid out during the first year of the contract.

Though Babel X isn’t by any means new to the Department of Justice’s toolkit, these 5,000 licenses are a significant increase in its investment in such social media surveillance. Keeping a close eye on general social media activity has become a priority for both the DOJ and the federal government writ large ever since the January 6 Capitol riot, which was fueled in no small part by organizing on social media.

Sure, preventing similar incidents from occurring in the future would be ideal. But sacrificing the privacy of the general public to do so isn’t the way to get that done.

Watching, well, everything — As part of its hunt for the perfect social media surveillance software, the FBI handed out a list of platforms it hopes to watch in real-time. Topping the list of required platforms: Twitter, Facebook, Instagram, YouTube, LinkedIn, the dark web, VK, and Telegram. That will cover a significant amount of internet activity, and not just from the U.S. VK, in particular, is most popular in Europe.

WaPo says the FBI also listed some platforms it would prefer to monitor but that weren’t strictly necessary, including the likes of Reddit, Discord, Snapchat, TikTok, Weibo, 8Kun, Gab, and Parler.

The FBI estimates its thousands of licenses will be able to run about 20,000 keyword searches every month. If there’s even a semblance of silver lining here, it’s that the FBI has only requested that publicly available information be searchable by its licenses; the Bureau says this process won’t be “intrusive” since it’s all available on the web anyway.

Experts agree: it’s bad — More troubling still is the way in which the FBI plans to utilize this new trove of data. The Bureau isn’t just collecting data — it’s attempting to interpret it. More specifically, the FBI wants to analyze social media patterns to identify bad actors — a process that essentially amounts to educated guessing.

Internet safety experts and activists agree that this process will be highly problematic. Speaking to the FBI’s hope for predicting the future, ACLU staff attorney Matt Cagle said: “… not only is this impossible, in all likelihood it will risk further bias and harm against the same people that the government has historically mislabeled as suspicious, including movement leaders, immigrants and members of religious and ethnic minorities.”

The FBI’s arguments in favor of surveillance are eerily reminiscent of those utilized by Clearview AI, which has scraped the public web to create a facial recognition database of 10 billion images. The use of Clearview’s software has been condemned by many lawmakers; whether or not they’ll feel comfortable being as vocal about the FBI’s new contract remains to be seen.