Internet filtering laws outdated before they are passed

15 February 2010

The Federal Government's plan to compel Internet Service Providers to filter the Internet has been unanimously opposed by Australasian Computer Science academics, says University of Sydney Associate Professor Bjorn Landfeldt.

The policy, announced in December last year, is designed to improve cyber safety for families but Associate Professor Landfeldt says it is highly ineffective and could negatively impact the National Broadband Network (NBN).

"Proposed amendments to the Broadcasting Services Act to introduce compulsory filtering is unlikely to exclude much of the unwanted content," he says. "It is inapplicable to many of the current methods of online content distribution and has the potential to restrict Internet bandwidth."

Associate Professor Landfeldt says the Computer Research and Education Association (CORE) - representing Australasian Computer Science lecturers and professors - asked him to publicise its opposition to the legislation at its recent Annual General Meeting.

"It's the first time I've seen this group so quickly reach consensus in its opposition to a government initiative," he says.

CORE instead asks the Government establish a working party be put together to properly attend to the many issues surrounding the legislative amendments.

"An alternative is to again make available the PC-based home filters the taxpayers paid for under the previous government, saving some of the $125.8 million earmarked to improve cyber safety in the 2008 budget," Associate Professor Landfeldt says.

CORE's full statement on internet filtering follows.

CORE press release

The Federal Government has announced its intention to introduce new legislation to compel Australian Internet Service Providers to filter all information transfer in Australia, with the intent of stopping the general public from accessing selected information.

CORE, the Australasian association of Computer Science schools and departments, has strong concerns about this policy. It is CORE's view that the proposed list-based filtering will not be effective and brings several risks. It is unlikely to exclude much of the unwanted content; it is inapplicable to many of the current methods of online content distribution; and it has the potential to restrict Internet bandwidth. For these reasons, we believe that a better approach is to form an expert body to help design and implement an approach or scheme that is appropriate for Australia.

The Government has decided against mandating dynamic content filtering at the ISP level, a move that CORE supports. Such filtering is only scalable if it is distributed on users' end-systems, and any attempt to centralise it would inevitably lead to performance penalties for both responsiveness and bandwidth.

A key concern is the limitations of list-based filtering schemes, which build on reporting by the general public and actioning by a Government-nominated organisation. With the pace and volume at which content is added to the Internet, such lists can only capture a small fraction of the material that would be classified as harmful. Also, the emergence of short-lived data such as live data streams and dynamic content generation, and the use of dynamic addressing, leads to a situation where any given list rapidly becomes inaccurate or obsolete. It is therefore unlikely that any significant protection can be offered by such an approach.

The filtering has been proposed as a means of making the Internet child-friendly or child-safe, and there is therefore a real risk that, once the filter is put in place, parents and teachers will become less vigilant in supervising children's on-line activities. This is despite the fact that methods for circumventing the filter are widely known. The Government has proposed to also make resources available to educate the general public about cyber safety, a move CORE fully supports. However, we feel that such education must stress the limited benefits of list-based filtering schemes in order to minimize the risk of lowered vigilance.

The ease at which a list-based filter can be defeated further limits the effectiveness of any such scheme.

CORE is concerned that the Government has not carried out any study to investigate the mandated filtering in the context of future networking environments. The pilot study during 2009 was designed to replicate well-known properties and no significant new information was presented, although the underlying technology is rapidly changing, in particular due to the National Broadband Network, NBN. Not only will the NBN bring higher data rates and shorter response times, it has the potential to overturn our current Internet habits and pave the way for a much richer multimedia experience. It seems certain that the simple web page with relatively unchanging information will decrease significantly in importance. In addition, user-generated live content and peer-to-peer networks will inevitably mean that the filtering scheme has to evolve and be extended to incorporate other, more resource-intensive filtering mechanisms. CORE therefore sees a tangible risk that the mandated filtering will pave the way for costly solutions that have a negative impact on the NBN.

In addition to the many technical concerns held by CORE, there has been wide reaction from other sections of the community to this form of censorship, around concerns with issues such as freedom of speech, the potential for misuse of the powerful tool of blanket filtering, or whether the scheme does indeed achieve the stated aims. CORE asks the Government to drop its plans for legislation of mandated content filtering and create a broad and inclusive working party consisting of experts from the public sector, industry, and academia who can together properly investigate the many issues associated with such a scheme before a decision is made to mandate its implementation.

CORE sees at best limited benefits from the proposed filtering scheme, and high potential risks if the scheme were to be implemented without proper investigation of the possible side effects.

On behalf of CORE

A/Prof. Bjorn Landfeldt

University of Sydney

Prof. Christopher Lueg

University of Tasmania

Contact: Jocelyn Prasad

Phone: 02 9114 1382