Picture the scene: a music-industry debate hosted in a small room above a pub in Soho, London. "Just answer me: Are you technically able to stop people sharing our music illegally online?" The casually dressed moderator's voice was tense. It was the third time he'd asked the question.

April 30, 2009

4 Min Read
Operators can't play dumb if they want to be smart pipes

By Rob Gallagher

Picture the scene: a music-industry debate hosted in a small room above a pub in Soho, London. “Just answer me: Are you technically able to stop people sharing our music illegally online?” The casually dressed moderator’s voice was tense. It was the third time he’d asked the question.

The telecoms executive nervously adjusted his tie and replied that ISPs were much like the postal service. They couldn’t possibly be held responsible for the content traveling over the networks.

That was nearly two years ago. I’ve taken some dramatic license: My notes don’t show whether the telecoms executive actually fidgeted with his tie. Maybe he paused to wipe a bead of sweat from his brow…. But you get the point. Here you had two people from two different worlds trying to get each other to understand the pressures their businesses were under.

At the time, both arguments were valid. Even then, ISPs had a variety of means at their disposal to block illegal activities on their networks. In fact, many had been covertly using so-called deep-packet-inspection (DPI) technology to identify and discriminate against peer-to-peer file-sharing for several years to reduce the high cost of supporting the traffic.

This revelation rubbed more than a few copyright holders the wrong way, especially given the ISPs’ insistence that they had little control over how their networks were used.

The ISPs countered that they were discriminating against all P2P traffic, not just the illegal kind. DPI or not, they were no more able to inspect every packet of data traveling over their networks than the postal service is able to open every letter or package.

ISPs also have the protection of the law. Under European regulation, for example, they are “mere conduits,” carriers of information rather than publishers.

Yet more and more operators seem to resent this dumb-pipe status. They instead want to become smart pipes by using DPI to delve deeper than ever before into what’s happening on their network.

Some hope they will be able to use this information to prioritize certain Internet services, and charge their providers for the privilege. Others also want to use DPI to help deliver targeted advertising using systems such as Phorm and NebuAd.

Both ideas are already controversial. A number of campaign groups are concerned that allowing deals to give certain services priority on their networks would tilt the level playing field of the Internet toward ISPs and wealthy content providers. The European Parliament and Council of Ministers plan to reduce this threat by pushing for legislation to allow national regulators to force ISPs to provide a minimum quality of service for open Internet traffic.

And earlier this month, the European Commission launched infringement proceedings against the UK after receiving several complaints from citizens worried about the privacy implications of BT’s trials of Phorm.

Plans to use DPI for profit also lend weight to copyright holders’ claims that ISPs can and should take more responsibility for stolen content traveling over networks. This could in turn encourage an array of interest groups to call for action on online activities they are unhappy about.

Whether this would result in a change in the law is unclear. What is clear is that copyright holders are taking a great interest in DPI and plan to use their knowledge to lobby policymakers. The question for operators is whether the commercial gains from deeper DPI will outweigh such consequences.

There seems to be little appetite among content and application providers to pay ISPs to enhance the quality of service of their offerings. Phorm’s controversial approach to advertising, meanwhile, could cost ISPs more in lost subscribers than new advertising revenues, if the company’s critics are to be believed.

And just because an ISP is using DPI for these purposes does not mean it can easily use the equipment to meet the needs of, say, copyright holders or the authorities.

Most use DPI to target generally heavy usage or prevent denial-of-service attacks and other security risks. Using the technology to identify actual sharing of illegal content would place a major burden on their systems, which could slow down the network, they claim. Content-recognition technology would help, but copyright holders would have to play a role in managing it.

It is this kind of knowledge that ISPs need to share with the wider world. DPI is by nature controversial, and shrouding it in secrecy will enflame its detractors. Unless ISPs are open and honest, the next time they face a grilling about the technology, it might be in a brightly lit government building rather than a cozy pub.

Read more about:

Discussion

You May Also Like