Mountain View, CA, United States (4E) – Publication of Google’s “Policy Violation Checker” patent application reveals the extent to which Google, and certainly other large companies, are using Big Data to automate spying on everything consumers do.
Inventors Mayank Talati, Dan Belov, Gary Young and Ashley Veselka, filed patent application US20130110748-A1 with the stated intent “to prevent violations of company policy or laws before they occur”. While this is noble idea it is also fraught with problems.
The patent application reveals an invention where a computer monitors every word of a communication to check for “problematic phrases”. All phrases in the communication are compared to a database of “problematic phrases” and if a match is found the user is notified. This shows us that Google and undoubtedly Bing/Microsoft, Yahoo and the NSA are able to track consumers down to a word-by-word level on every communication they have access to.
A Computer Security Consultant (we will call Thomas) who asked to not be named said, “The patent application is scary. It basically enables a few people to control a large population in a very personal, automated and one-on-one way. Who designs the database of ‘problematic phrases’, who designs the user alerts and how the data about the users is distributed is too much power not to abuse. It is similar to Batman’s cell phone surveillance system shown in The Dark Knight, only it does this with emails.”
The concern here is a system like this can give too much information about an individual’s personal life and that that information can be abused. A “problematic phrase” database rife with phrases designed to pin down specific groups of people could enable mass and automatic discrimination. Phrases that include the names of Gay bars, supplements that purportedly cure specific ailments or political affiliations could trigger what appears to be a friendly alert to user, but then add notes to their profile so they can be later discriminated against.
Another problem with this system is that it does not allow for context to determine the actual meaning of the event. Just like a red light camera will give you a ticket if you are speeding to the hospital on an empty road these “problematic phrases” can’t tell if you will are will not “kill” your sister in-law for not bringing her famous chocolate-chip cookies to the party.
While the patent does simply end the technologies stated use at notifying the user it is clear that is system can do much much more.
“But this genie will never go back in the bottle. The patent application does not make this type of spying go away it only gives one company, if the patent issues, ownership over it. And the NSA is not even bound by patents. You’d better watch what you email from now on.”, said Thomas.
Claim 6 further adds to users angst. It describes how the invention can “[alert] a third party to a match between a textual phrase and a phrase in the database having legal implications or violating policy.” This means that simply typing something into your Gmail editor, before sending the email, could get you a call from HR, your corporate legal department or the FBI.
If the patent does or does not issue will not stop this technology for being used. Only legislation and enforcement can do that. We will see the full impact of this technology in 10 to 15 years. Sooner if we are vigilant in helping keep it honest.