AT&T Investigating Copyrighted Material Network Filter
In one of the more interesting stories to come out of the telecom industry in recent weeks, the New York Times is reporting that AT&T is currently in the testing phases of developing a network filter that will monitor their networks for copyrighted content. The company is reportedly in talks with content owners such as NBC Universal regarding incentives for the company to filter copyrighted material.
This brings up a host of interesting security related issues regarding the role of network carriers, their legal responsibilities, and the privacy of their end users. The most obvious of such concerns is that ISPs (plural if we consider other related cases such as Comast) are moving toward the active monitoring of the content flowing over their networks. While in the past they have simply provided the pipelines and turned a blind eye to what actually went through them, it appears that may no longer be true.
As so often seems to be the case with emerging technologies, it isn’t the research of the technology that is most concerning, but rather the logical path down which such a technology will likely travel. In this case, that path seems to point to ISPs restricting what can and cannot be sent over their networks. Not only does such a scheme seem destined to present a host of practical problems (such as correctly identifying what is, and is not, legitimately transmitted content), but it also implies detailed logging of customer network activity. In order for networks to determine whether content is in violation of copyright, they will have to gather a significant amount of information about that content and its ownership. This stands in stark contrast to Comcast’s practice of relatively blindly blocking a specific protocol, as it suggests an awareness of the actual content itself, and not merely its mode of transmission (bittorent). This is a particularly important distinction, as it is the difference between looking at envelopes in someone’s mailbox and opening those letters and reading their contents: one is far more invasive than another.
What is also worrisome is the potential loss of any remaining sense of anonymity on the Internet. Though increasingly difficult to do, one could argue it is still possible to browse the internet over an unencrypted connection with relative privacy if the proper measures are taken. If ISPs begin monitoring individual connections for content, the number of proxy servers used or IP masks employed will be meaningless when logging of one’s online activity takes place not at the end server being contacted but at the end of their driveway. What isn’t clear is where in the network content monitoring might take place, and whether traffic that neither originates nor terminates on AT&T’s network but instead merely traverses it will also be affected. If so, the impact of the program could be especially large, and those Internet users who aren’t AT&T customers will have little recourse as they will have no control over their data beyond their own ISP.
The question that begs to be asked then is what will be done if and when copyrighted material is detected on the network, and what will it mean for the integrity of the customer’s data. AT&T has said it will not simply drop offending network traffic, yet surely an investment in monitoring technologies on this scale is not merely for the sake of better traffic characterization, but rather implies an intent to take action when content is flagged. It is hard to imagine a solution that does not curtail the availability of the end users’ data to do with as they please.
In the end, one has to wonder what economic forces must be at play to make AT&T consider taking these measures. Not only will it open up a potentially massive legal liability for the content flowing over their networks (no longer will they be unaware and thus not not responsible), but clearly, based on the reaction customers gave Comcast when news of their networking monitoring leaked, the move will likely be hugely unpopular. Considering that AT&T must be well aware of such considerations and is still willing to pay to develop the technology, we can only conclude the gains it expects from lower bandwidth utilization and incentives from its media partners must be extensive. Unfortunately, while such measures will likely be a boon for intellectual property rights, it seems equally likely that they will have an especially negative impact on the availability and confidentiality of customers’ data.