The government's announcement yesterday that it would press ahead with its mandatory ISP filtering scheme is as unsurprising as it is disappointing. Many people with concerns about the plan had hoped that the Minister would finally get some good advice, get distracted with other things, or otherwise come to his senses and take a face-saving out. Sadly for Australia that was not to be. Yesterday, the Minister released the report detailing the technical results of the trial, and at the same time announced the Government was pushing ahead with the filter, with legislation to be introduced next year. (Continue reading at ABC's Unleashed.)


  1. Interestingly it seems Conroy claims 100% success of the trial even though no target to be reached was setup before the trail. Without target you can't have 100% success, oh wait you can if you're in Australian government.

    Comment by Stefan on 17 December 2009 at 20:48
  2. For clarity, the target was whether the filtering systems used actually blocked the URLs on the test list. This was a requirement of the filtering systems used. You can read the full study here where they repeat that over and over:

    The document states that:

    Comment by trog on 17 December 2009 at 21:57
  3. Woops, looks like the blog ate the rest of my post - here's the last part:

    The document states that: For the purposes of the pilot it was necessary for each participating ISP to achieve 100 percent blocking of these sites before any further testing was conducted.

    So the target was reached, but it was such a low-hanging, trivially accessible target that it's almost meaningless. Further, it accomplishes nothing at expense and inconvenience that will be passed on to the Australian Internet user, not just in taxes to pay for the government side of this silly scheme, but in the increases costs ISPs will have to absorb to enforce it.

    It's almost entirely unsurprising that they got a 100 percent blocking rate, because it was clear they wanted to do further testing, but anyway.

    I've only glanced through the document, but there's no real definition over what it means to block a site. I suspect they just mean that under normal circumstances, a user typing into their browser is not allowed to access it.

    This is a trivial technical goal to achieve - if you don't care about performance or cost. Important to remember.

    Enforcing it rigidly is another measure. There's a section of the PDF which deals with circumvention, where the enforcement was much less than 100%, ranging from 8.1% successfully blocked at the low end, to 94.5% at the high end, although it's noted that the high end had 'noticeable performance degradation'.

    No details about the circumvention methods are provided, 'for public interest reasons', but I'm guessing that this information is not yet illegal to possess or distribute so I look forward to doing whatever I can to make sure people are aware of secure circumvention techniques for the inevitable occurrence of when something is blocked that should not have been.

    Comment by trog on 17 December 2009 at 21:59
  4. Conroy's press release touting '100% success' was deliberately misleading. This figure relates to *known sites* on the blacklist only.

    It's the automated content scanning that's hard and that slows down the internet and causes overblocking. These filters only achieved 78-84 percent accuracy according to page 2 of the report, and they overblocked blocked 3.4% of 'safe' content as well. But those numbers didn't make it into the press release. This is deliberate misdirection by the minister and unfortunately the media have taken the bait.

    I'm absolutely disgusted that an Australian minister would attempt to mislead the public on any matter related to censorship. Its intolerable behaviour in a supposedly-free country, and I sincerely hope his electorate hands him his arse in the next election.

    Comment by Simon on 23 December 2009 at 05:04