Press release: iWeek Opinion: Intermediary liability and the Internet: don’t shoot the messenger
Peter Coroneos, Chief Executive, Internet Industry Association (IIA) of Australia
The Internet has developed in an anarchic way since it escaped from ARPANET, with much of the growth driven by the intermediaries that deliver broadband Internet services as well as the compelling content and applications that take advantage of high-speed Internet access.
Search engines, social networking sites and Internet Access Providers (IAPs) are all helping to fuel the growth of the Internet around the world. Collectively, these Internet service providers have one thing in common: they have limited control over what their users do once they’re online.
Nonetheless, they’re all increasingly finding themselves in the crossfire between users on the one side and governments and old economy companies that are trying to impose control on the Web on the other.
The upshot is that nearly anyone with a gripe about how the Internet is used is trying to place the responsibility of control on IAPs, social networking sites, and other intermediaries. The tension is playing out in legal debates around the world.
In Australia, for example, the new Labour government aims to introduce measures to require all Australian IAPs to implement server-based filtering systems to block access to illegal and other inappropriate material subject to a blacklist compiled by a government agency.
The policy would require IAPs to deliver ‘clean’ feeds to households and schools. The big question for IAPs and other intermediaries is whether it is desirable to be positioned as the Net Police. If so, is it even possible for industry or the government to enforce the restrictions they want to put in place?
In response to the first question, IAPs around the world are helping governments to address problems such as child pornography because it’s becoming harder for them to distance themselves from the debate. Once a government decides to implement a policy such as Australia’s ‘clean feed,’ it becomes difficult for IAPs to refuse to cooperate since they could face more onerous regulations in future if they don’t play ball.
The question of whether IAPs can effectively regulate the content that their users can access is more uncertain in the world of Web 2.0 and peer-to-peer applications. One problem lies in the fact that filtering software is still far from perfect.
It’s not uncommon for a content filtering solution to allow material that should be forbidden to sneak through, yet accidentally block harmless, legitimate content. This is not an issue where filters are used voluntarily by users and can be turned on or off, but it is problematic where filtering occurs by government mandate and the lists are not transparent to the public.
In countries where there is no constitutionally guaranteed right of free speech such as Australia (which has only an implied right to political discourse), the question of transparency becomes even more fundamental. There are possibilities for abuse and ‘scope creep’ which could see block lists expanded over time according to the political whims of the day.
IAPs and service providers generally are also concerned about the precedent a mandatory filtering approach creates in areas like copyright infringement, defamation, pharmaceuticals, and public health. In any case, determined users will usually find a workaround that will allow them to access content that governments would prefer them not to see. We see this in totalitarian states with the strictest laws.
Then there is the issue of who should bear the costs of installing and managing expensive and complex filtering software.
A 2004 Australian Government review into the filtering technology found that IAPs face implementation costs of around AU$ 45 million (about R297 million) and ongoing costs of more than AU$ 33 million (about R217 million) a year to put in place a technology of questionable benefit. Costs are sure to be higher by now.
The final costs will depend largely on the outcome of technical “real world” trials slated for later this year. As well as providing indicative costs, the trials are intended to quantify the effect on network performance. Only then will we see what is involved.
Another concern is that filters applied at the IAP server level may degrade internet access speeds. Recent studies indicate that filters may produce 22% to 80% degradation in access speeds; the best performing filter in a recent lab test showed a latency of 2%, but had the highest error rate.
The Australian Government has, however, indicated that it will not implement a policy that will have a substantial negative effect on access speeds so IAPs are hopeful the policy can be limited to child pornography sites.
When it comes to protecting children from dubious content on the Internet, there is no substitute for responsible adult supervision. While there will always be a place for appropriate regulation, education and end user empowerment supported by international cooperation at industry and government levels will provide the best outcomes in the end.
For further information, please contact the ISPA secretariat on the Contact ISPA page.