Has Telstra made 84,000 reasons to filter?

Has Telstra made 84,000 reasons to filter?

Summary: The heated debate about internet filtering has calmed down this year, with voluntary filtering at several ISPs drawing little of the chanting, marching and petitioning that we saw two years ago. Does this fact, paired with recent statistics showing that Telstra's filter is keeping at least a few sickos from casual child porn access, legitimise the case for moderate internet filtering?


Like bellbottoms, herpes and anything to do with the Kardashians, the spectre of internet filtering just never seems to go away; it lurks in the background for a time, then explodes back to the forefront just when we'd managed to forget about it and move on to other things.

Forgetting about the filter has been relatively easy, recently; with the government officially in a hiatus while the largely ignored classification review lumbers along, the NBN has been more than enough to keep Stephen Conroy and his lot quite busy. The only real progress on filtering this year has been the decision by ISPs Telstra, Optus and CyberOne to implement filters blocking a worst-of-the-worst list maintained and securely distributed around the world by Interpol.

Protesters have painted filters as being the Devil's work, but with Telstra blocking more than one child porn visit every minute, what if we really do need it?
(Protest Series 1 images by Gary Cowles, royalty free)

By all accounts, it's working; Telstra's filter, we learned recently, blocked over 84,000 requests for child porn sites in its first three months of operation. It's not clear how those requests relate to multiple queries by pages with multiple elements, or how many individual users they account for; they could be from just five or ten porn-obsessed customers, or they could represent a broader demographic within Telstra's user base.

That this figure was released at all, however, is an important milestone in the filter debate. We know filtering works, it does not appear to have brought Telstra's internet to a standstill and there is a demand for it within Australia's population. Yet, with two other ISPs currently filtering, and neither prepared to share their filtering numbers or compelled to do so, Telstra's number raises bigger questions about how Australia's filtering regime will develop.

Will the eventual filtering environment be based on open and transparent blocking as its grudging advocates demand? Or — as Australians would most definitely resent — will it be an opaque, secretive filtering with no real understanding of how or whether it's working at all? This sort of opacity, after all, was a major tier of Electronic Frontiers Foundation's protests against the ACMA-administered filter.

We know filtering works, it does not appear to have brought Telstra's internet to a standstill and there is a demand for it within Australia's population.

Certainly, ACMA's experience in 2009 showed how not to do a blacklist. With an internationally accepted standard for filtering, however, arguments about the subjectivity or political control of a domestic filter would seem to be diminished somewhat. Filtering, it seems these days, may well be bound to happen one way or another; just like Telstra's separation or telcos' last chance to improve their less-than-optimal customer service, ISPs are simply being given the chance to choose the manner in which filtering happens.

Is there still value in resisting the filter on philosophical terms? Perhaps. But the AIIA is on record saying that most ISPs will be filtering based on the Interpol list before year's end — and they're going to have to decide how much they share about their experiences. Not to do so would test the high water mark of hypocrisy, since ISPs once railed against the lack of ACMA visibility.

So why haven't Optus and CyberOne followed Telstra's lead by sharing their filtering numbers? And would other ISPs share theirs? There are several possibilities.

First, of course, is the chance that Optus and CyberOne, Telstra's fellow voluntary filterers, actually aren't collecting statistics on utilisation of their filters. Yet this would seem to be absurd, not the least because any ISP, and any filtering platform worth its salt, would have no trouble generating logs to track exactly these kinds of statistics.

There's also the possibility that they are collecting the data, but don't want to publicise it out of fear of giving either side of the filtering debate ammunition with which to argue their positions. Say it was revealed that the three ISPs saw over 150,000 requests for child porn materials in three months — the equivalent of 1667 requests per day, or more than one per minute, just by customers of those three ISPs — a case could certainly be made that broader filtering is perhaps not as crazy as opponents initially feared.

Yet, I suspect a third dynamic is in play here: ISPs are looking for solid footing in a rapidly changing legal landscape that extends far beyond child porn, and they don't want to load the decision-making process with new talking points before they get some certainty in areas such as data retention and copyright enforcement. ISPs like Exetel, TPG and Internode have already indicated that they'll implement filtering when they have to — but, for now, they have other priorities.

iiNet's High Court hearing against AFACT in four weeks, for example, will be closely watched as a marker of the extent to which ISPs must intervene with customers' online activities. Ditto New Zealand's new three-strikes law, which has already seen more than 1000 warning notices sent out since it went into effect on 1 September.

ISPs are looking for solid footing in a rapidly changing legal landscape that extends far beyond child porn, and they don't want to load the decision-making process with new talking points before they get some certainty in areas such as data retention and copyright enforcement.

If Australia's courts make a case that ISPs must be more proactive about monitoring customers' activities, it's not a big philosophical step to extend that to general internet usage and the filtering regime that Labor pushed so hard for. But will ISPs avoid potential liability by taking a don't-ask-don't-tell approach for as long as possible? Possibly. But should they? Certainly, those reflexively emotional proponents of the #nocleanfeed debate would advocate that ISPs put off filters for as long as possible.

Yet, in viewing the volunteer-filtering experiment with cool objectivity, we must consider whether something fundamentally valuable has been accomplished. We're not talking about ten requests for what is universally held to be morally reprehensible content, or a hundred; Telstra's figures alone suggest 28,000 requests for such material every month from its user base, which represents nearly half of all Australian internet users.

For those that rightly questioned the difficult and fault-prone methodology employed by ACMA in building its contentious list, the idea of supporting filtering will be as appealing as eating handfuls of glass shards and razor blades. Yet there is something seemingly less offensive about having an internationally recognised police agency, which imposes checks and balances on its blacklist and operates outside the feared political influence of Australia's ever-changing political mores, determining the make-up of the blacklist.

Australia isn't going to stop the world's child porn scourge single-handedly, but a filtering policy done right — and transparently — may ultimately turn out to be an acceptable compromise between Labor's central filter pie in the sky, and the every-man-for-himself libertarianism of the anti-filter brigade. If you accept censorship of the worst-of-the-worst as a good thing, and accept also that the nature of that censorship is as even-handed as you're going to get in today's world, then perhaps you will accept that the type of voluntary filtering being undertaken should be adopted by other ISPs as well.

Good policy is never born of extremism, but of cold, hard facts and solutions to improve them. So far, the filtering debate has been stoked by justifiable outrage, and desperately short of real-world data; therefore, one would hope that other ISPs implementing filtering will share the results so that we can collectively judge whether a moderate level of filtering is a good idea after all — or just an irrevocable step towards the authoritarian state that so many have feared.

What did you think of Telstra's filtering numbers? Does it support the case for filtering to Interpol's list? Is the voluntary solution preferable to Labor's government-imposed, domestically enforced alternative? And will the community still muster strong opposition to whatever filters are introduced?

Topics: Government, Broadband, Government AU, Telcos, Optus, Telstra, NBN, IT Employment


Australia’s first-world economy relies on first-rate IT and telecommunications innovation. David Braue, an award-winning IT journalist and former Macworld editor, covers its challenges, successes and lessons learned as it uses ICT to assert its leadership in the developing Asia-Pacific region – and strengthen its reputation on the world stage.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • I don't think anyone has a problem with this, but it is troubling that this same filter was expanded last week in the UK to include a non-porn site offering movie downloads. If this kind of functionality creep happens, what is to stop AFACT demanding that every torrent, streaming music service, VPN provider, non-accredited reseller be pulled from the internet?
    • A lot of people demanded that any filter based decisions should come from a court order, the UK decision was a court order. Fully tranparent, open court proceedings, with full appeal rights. AFACT and other groups will never be able to prove that every torrent, streaming music service etc etc was produced and is being used predominantly for an identified illegal purpose. They could prove it in that court of law for the website named in that case.
  • Hold on there David...just what exactly is WRONG with bellbottoms??? I still have 5 pair (in case they come back again...)!
  • There are a couple of things about the filtering that you have talked about. Firstly the "voluntary" filter that BigPond is using is a Domain Name filter. The Labor mandatory ISP filter is a URL filter. Both are very easy to circumvent. Even Senator Conroy has admitted to Senate Estimates that a Domain Name filter is likely to over block. This means that a perfectly legal web sites could be blocked because they share a domain with a site that contains "worst of the worst" material. Interpol's own web site admits that this is possible.

    The so called Telstra (should be BigPond surely) figures are really pretty meaningless all they tell us is how many requests have been made to access a domain and been referred to the Interpol Stop page. We don't know how many are search engine crawlers, how many are repeat requests and how many were genuine attempts to access child abuse material. The number of sites on the Interpol list according to an Interpol operative when questioned in one of the European Parliaments is less than 300. 84,000 requests to access less than 300 very dodgy sites in less than 3 months seems a bit suspect to me.

    Finally I would query whether the basic premise that filtering of the "worst of the worst" protects children is correct. In the abstract for Pornography and sex crimes in the Czech Republic. Diamond M, Jozifkova E, Weiss P. The following is quoted "Of particular note is that this country, like Denmark and Japan, had a prolonged interval during which possession of child pornography was not illegal and, like those other countries, showed a significant decrease in the incidence of child sex abuse."

    It seems highly possible that the filtering of the internet using either method could in fact do more harm than good.
  • I think that one should be very cautious about equating filter "hits" with genuine attempts to access child pornography.

    The Telstra filter is domain-based, not URL-based. This means that anyone attempting to access any content anywhere on that domain, regardless on whether it is child pornography or not, will register a "hit" with the filter.

    For example, the reason why a Queensland dentist's website appeared on the leaked ACMA blacklist was allegedly because some hacker had placed child pornography in a hidden sub-directory on the site. Anybody with a toothache visiting the URL http://qlddentist.com.au/ [not the real site] would see the regular dentist's page with non-offending content. Only those visiting http://qlddentist.com.au/secret-content/ will actually be served with child pornography. However, with domain-based filtering, both URLs will be blocked, and both URLs will register "hits" with the filter.

    IMO it's unlikely a significant portion of the 84000 figure are genuine attempts to access child pornography. The numbers just don't add up. However, without more detailed logs and a more careful analysis, it is impossible to say for sure. Unfortunately I think it is unlikely Telstra will release more information beyond scary sounding figures.
  • What few comment on is that the filter can be bypassed easily with VPN so why bother?
    Hard core child porn customers probably already use this.

    Half the hits reported could even have been people testing the filter out of curiosity.
    • Half the people knew what sites are on the Interpol list and went there, exposing themselves to potential AFP tracking, just out curiosity? Doubtful.
  • I am thoroughly sick and tired of hearing about porn being the reason for the filter. It is not! Get over it people!

    It's about information control, pure and simple. The fact that the blacklist will be managed by Interpol, an impenetrable and unaccountable outreach of the UN, should sound alarm bells for anyone with half a brain.

    Be assured that while we've been sleeping, the web-bots have been busy seeking out every page with politically incorrect and/or embarrassing content, ranging from questions about the official 9/11 story (e.g. the WTC7 weirdness), to the real reason for the Kennedy assassination, to refutations of global warming, to information which suggests the deliberate toxificatin of our food, water and vaccines, to questions about the official holocaust story... and so on, and so on...

    Once the filter cuts in, don't be surprised when a massive number of websites promptly vanish from view, and the whole Internet gets dumbed down to politically acceptable newspeak.

    VPN freedom? Short-term perhaps, until the powers driving this madness find a way around that as well. Don't underestimate what they are able to do.

    Goodbye freedom of thought, and hello brave new world!