AT&T’s latest patented method for real time detection of files deemed to infringe copyright seems to have stuck in the throat of libertarians concerned that this really will spell the end of Internet freedom as they know it. This is probably because it has coincided with all the revelations about state and commercial sponsored intrusions of piracy via the Internet, especially the ongoing Edward Snowden affair. It is hard to see why else it should evoke such concern, although AT&T itself will be grateful for the publicity and hopeful that its patent will instill an equal chill among copyright infringers.
Perhaps it has offended libertarians because its target is not organized piracy but the emerging threat of unauthorized Internet redistribution of content by individuals who may be oblivious that they may be causing offence. The patent itself, which we have inspected via the references cited in the document, breaks no new ground and it is the proposed manner of its application that is most interesting. The underlying algorithms are derived from several papers published around a decade ago on defining and identifying anomalies in traffic flows across the Internet. What AT&T has done is incorporate these traffic flow analysis methods into algorithms that can run in real time and identify any file that fulfils one of a set of criteria defining copyrighted material. The rules would recognize defining characteristics of a file and the patent states that it is preferable to use adaptive analysis allowing the system to extract its own signatures, effectively applying machine learning to perfect the recognition and virtually eliminate false negatives or positives.
There is no reason to believe AT&T has stolen a particular lead over competitors, especially as it has been collaborating with some of them under various initiatives. For example in 2008 it teamed up with Microsoft, Cisco, Viacom, and NBC Universal to create a lobbying coalition named “Arts & Labs,” targeting piracy. Then in 2011 the MPAA (Motion Picture Association of America) and RIAA (Recording Industry Association of America) signed up with AT&T, Cablevision, Comcast, Time Warner Cable, and Verizon to create the Center for Copyright Information (CCI) in the US. The CCI’s terms allow ISPs to warn and punish subscribers for breaking the law and then to inform copyright holders about repeat offenders.
Debate has focused on how such techniques will be applied now by AT&T and others. AT&T stated that the new method is aimed at tackling the increasing and recurring problem of unauthorized content retransmission over the Internet. ‘Policing such unauthorized transmission is difficult for copyright owners, because the sources of copyrighted materials may be elusive, or indeed may be legitimate possessors of copyrighted materials, but do not have authorization to permit copies to be made,’ AT&T stated. ‘Pursuing the illegal distributors of such materials is problematic because the users are often numerous and diffuse and individual legal action against multiple small users is expensive ‘ as well as unsympathetic from a public relations standpoint when the users turn out to be teenagers or others whose motives are seldom to make a criminal profit.’
AT&T says the invention can score not just by identifying likely incidents of illicit content transmission, such as piracy of copyrighted material, but also confirming the presence of such content in the first place.
It is also designed, AT&T claimed, to preserve the privacy of those ISP customers who have no association with copyright infringement, although we have to say it is not clear how the patent achieves this. That is surely more a matter of intent than technology, given that the mechanism is supposed to be capable of detecting infringers and if necessary reporting them to copyright owners or law enforcement agencies. In fact it is precisely that aspect of the technology that alarms privacy advocates. The writing was put on the wall here by the charter of the CCI formed last year, which empowered ISPs to warn and punish subscribers for breaking the law. Each ISP in this club is bound to inform copyright holders about repeat offenders. If a court grants a subpoena to the MPAA and RIAA, then personal details can be handed over.
AT&T’s new algorithm provides the teeth for executing the CCI’s charter. The Telco, and probably all members of CCI, appear to be considering a series of escalating responses to people caught infringing copyright by the system. It is clearly impractical, counterproductive and heavy handed, to weigh down heavily on first time offenders, so the idea is to start with softer warnings.
Leaked documents indicated that AT&T at least may be considering a kind of ‘six strikes and you’re out’ policy. It could start with a gentle reminder or warning email, perhaps followed after a second offence with a requirement to take an online copyright course, rather as people caught speeding in some countries have the option of sitting a course and being told how dangerous their behavior is in lieu of having penalty points on their driving license.
But repeat offenders could be kicked off the service and even black listed, although that could well fall foul of civil liberty and privacy laws in a number of countries. The same could apply to AT&T turning over personal information to rights holders. Therefore it is a fair bet remedial action will be confined to shaming offenders and putting them through the grill through these online courses, reserving more threatening action only for the most serious offenders intent on making money through piracy. In the end, sensing an offence in real time and shutting down the connection temporarily is the Holy Grail and this is a step towards it.