Since the beginning of the current privacy scandal, Twitter has been careful to brand itself as a champion of privacy rights – but are they? As other tech titans first denied complicity, then joined together seeking permission to discuss it in the compliant American media, Twitter remained outside the fray.

In response to widespread public anxiety that Americans' social networking information is being funneled into the NSA's controversial PRISM program, a concerted effort has been made by Facebook, Google and Microsoft to be seen as defenders of privacy.

Microsoft released a statement on Wednesday urging the government to consider that "greater transparency on the aggregate volume and scope of national security requests, including FISA orders, would help the community understand and debate these important issues." Facebook made a similar public statement earlier this week, and Google asked the government "to help make it possible for Google to publish in our transparency report aggregate numbers of national security requests."

These statements were all deemed necessary by the social networking industry due to public speculation that these companies knew about PRISM or operations related to it, and they had to protect their reputations as guardians of private information or risk damaging their business. However, only a few members of the international press have begun to ask a new question: What about Twitter?

Le Monde asked whether Twitter's conspicuous absence from the controversy so far was due to a "defense of freedom or lack of interest?" Meanwhile, the most compelling version of this curiosity appeared when Chase Meacham at Policymic asked "Does Twitter just have better lawyers? Did the NSA pass it up? Or are they simply involved elsewhere?"

Twitter certainly has a better economy of trust than the corporations associated with PRISM's data-mining. For one, most tweets are in the public domain already, except for personal messages and deleted tweets. Furthermore, Twitter's legal team has been much more proactive than its Facebook and Google counterparts in protecting the privacy of its customers. They fought for two months against New York state prosecutors' attempts to subpoena the deleted tweets of journalist and Occupy protester Malcolm Harris before being forced by Judge Matthew Sciarrino to submit them, while in France the company mounted a similar defense of privacy in a court case.

However, the more complicated matter of deleted tweets – specifically, who owns and can access them – has important implications both for the Harris case and for Twitter's wider reputation as a defender of privacy. Twitter's policy of protecting customers' data is based on the legal stance that they don't claim to own it themselves, which is often misinterpreted as the customer therefore owning the rights to the data.

In reality, Twitter sells its database, named Firehose, to a very small group of companies: Gnip and Attensity, both based in America, and Datasift in the UK. Gnip, which supplies Firehose data to the Library of Congress, collects all the tweets on an hourly basis and then dumps that into the Library's internal database once a day, a process which suggests that Attensity and Datasift could have a similar capacity for data gathering.

Both Datasift and Attensity do not list all their clients, but according to Defense Weekly, Datasift sells twitter user information to Booz Allen Hamilton, the company at the center of the greatest domestic spying scandal of a generation.

The prosecutor in the Harris case may have been trying to set a legal precedent. He also may not have possessed the security clearance to know that Twitter was indirectly supplying all its data to American intelligence. He also may have been too miserly to purchase Harris's tweets directly from Datasift, who sell them at $1 per thousand. More likely, he was a knowing participant in a deception operation to make Twitter appear to be defending privacy in order to keep the world tweeting.

Booz Allen's access to DataSift's trove begins a process by which twitter feeds be funneled into various other data integration and visualization programs. One such tool is called MetaCarta, which can link documents to place names very accurately and weight the relevance of said documents. An individual tweet contains over 50 different metadata points, including the sender’s number of followers, source device, actual physical geolocation, and reported location. A picture of who is saying what about something and where can then be mapped very accurately.

MetaCarta was started with a grant from DARPA and received additional funding from In-Q-Tel, the CIA's proprietary non-profit venture capital fund that also financed the development of PRISM. One of MetaCarta's founders, Erik Rauch, drowned while hiking alone in the Sequoia national forest in 2005.

Attensity, Another In-Q-Tel financed company, fuses Twitter's Firehose database with other social media providers including Google+, Reddit, FaceBook, YouTube and various online reviews. It then extracts sentiment and opinion from full text searches. These sentiments can be tagged with names, places, products and linked to build a picture from social media to distill a picture of social sentiment.

A third In-Q-Tel financed company, Recorded Future, extracts time data from social media, including text searches, allowing the observer to know which tweets refer to not only to them, but to them on Tuesday, or next week or next month. Recorded Future offers its products not only to government intelligence entities, but to private companies to help them predict protests and public pressure campaigns. Their website's graphic highlighting this feature was taken at an Occupy Wall Street protest.

Recorded Future did a case study of tweets from the Libyan uprising to explain their technology. It states on their website: “Some people tweeted more as time went along, some fell off. Some people started making more predictions over time, some made less. Recorded Future lets the analyst quickly visualize both, as well of course allowing them to drill down to the actual underlying data.”

Fusing metadata across multiple visualization tools allows the government and corporate secret police forces to not only spy on people, but understand their sentiments and aggregate them across time and space. These are useful tools to crush dissent, or to channel and encapsulate it.

As an offensive cyberweapon, these tools allow the CIA to better influence, control and assist hostile political movements in other countries. These tools can be used to hatch, aid and launch coup attempts against unfriendly regimes abroad. American intelligence can now know where the greatest density of sentiment is both for and against consumer values, who is acting on those sentiments, and who is likely to act in the future.

Twitter has thus far protected its brand with help from intelligence agencies precisely because of its weaponized value against civil movements at home and for coup plotters abroad.


The Free Press welcomes London-based Lawrence Richards as our new foreign correspondent. Lawrence is a student at Goldsmith's College, University of London and the proud human companion of a pit-bull named Ginger.