Facebook Aims To Weed Out Piracy By Analyzing Your Personal Data - Insights

Facebook Aims To Weed Out Piracy By Analyzing Your Personal Data


3 minute read

Facebook has been granted a patent over the use of personal data -such as a user’s location, friends and interests – to identify potential piracy or other unauthorized content.

The patent, first unearthed by Torrentfreak, is titled ‘Using social signals to identify unauthorized content on a social networking system’. It’s not clear whether the system is already in use, or whether the company has firm plans to introduce it; Facebook’s promised us more information, but has yet to deliver.

But the patent allows for posts which are flagged by various algorithms to be disabled on the basis of both the information provided voluntarily by users – location, Likes and so on – and Facebook’s own data on its users.

“In addition to the information that users provide voluntarily, the social networking system can also record and generate additional information about its users, such as their communications with other users, how frequently they view certain pages or user profiles, their client device information, and their interactions with applications that operate on the social networking system,” the patent reads.

“The social networking system can also record the user’s browsing activity on external websites as long as those websites were accessed through links posted on the network.”

The application points out that while many users avoid listing their favorite movies and television shows on the site, it’s possible to deduce these from their activities, such as the external websites they’ve accessed – something it says hasn’t been done before.

Facebook has a record of cooperating with content owners and removing infringing content on the basis of complaints. Indeed, it’s been a little over-zealous on occasion, taking down perfectly legitimate pages including those of tech blogs Ars Technica and Neowin.

It’s not hard to imagine more such mistakes if this automated system is ever introduced.

“Basically, there’s always going to be the potential for real problems when legal or quasi-legal decisions are being made by automated systems,” says Sherwin Siy, vice president of legal affairs for campaign group Public Knowledge.

“If the patented invention (and we can leave for later whether it’d be a validly granted one) is being used, what are the consequences when it flags a file as ‘pirated’? What if it does so in error? Does that lock out a user? What recourse do they have?”

He adds: “Even if it’s just used to flag potential problems behind the scenes, if it’s used without users knowing, it can create a stigma on their accounts that could be invisible to them and impossible to correct.”

Intriguingly, while the application focuses almost entirely on copyright alone, the patent notes that the process could also be used to predict other types of unauthorized content. But Facebook also prohibits, for example, “organizations with a record of terrorist or violent criminal activity”, pornography and spam – all of which could be much easier to identify using these techniques than copyright infringement. Guess it’s just a case of priorities.