ISPs don’t have to collect voluntary filter data

10

news The Australian Federal Police has confirmed there is no obligation by Internet service providers participating in its voluntary Internet filtering trial to collect data about how many requests to visit the sites they block under the trial — or to disclose that data if it is actually collected.

The voluntary filtering initiative is a stop-gap measure agreed to by ISPs and the Federal Government in mid-2010 while a review is carried out into the Refused Classification category of content which Government’s wider mandatory filter project is slated to block. It has initially been implemented by Telstra, Optus, and a smaller ISP known as CyberOne and blocks a list of ‘worst of the worst’ of child pornography sites as supplied by international policing agency Interpol.

Customers who visit one of the sites on Interpol’s list are greeted by al ‘stop page’ which will explain that the content they have attempted to access is illegal, along with instructions as to how they can challenge Interpol’s ruling. Those who believe their web site has been inadvertently blocked by Interpol are able to ask for a review via the agency’s own website, or will be able to contact the AFP.

Last week the AFP revealed Telstra had blocked more than 84,000 requests to access blocked sites on Interpol’s list since the start of July this year. However, in response to questions on the matter this week, the AFP confirmed a report by iTNews which quoted the organisation as stating that ISPs did not actually have to provide data under the trial.

“Yes, the AFP can confirm that the provision of statistical data by ISPs is not compulsory,” a spokesperson said in a statement. “It is a decision for each individual ISP as to whether they provide statistical data. “Telstra is the only ISP that has provided the AFP with statistical information in relation to the amount of redirections that have occurred. ISPs are asked to provide non identifying statistical data when joining the trial; however the decision to provide the statistics remains with the ISP.”

They added the goal of the Interpol list was not to generate cases or investigations into Internet users who had attempted to access the blocked information; rather the main aim of the use of the blacklist was to protect the rights of the children depicted in the offensive material, by preventing further distribution.

A review of the trial is slated to be carried out later this year. Asked whether the AFP believed the provision of statistical data would be useful in judging whether the voluntary Internet filter was a useful tool, the AFP spokesperson said it was “difficult to judge” how successful the trial would be in blocking online access to child pornography. “Statistics are only one of the many measures which will be utilised to gauge the trial’s effectiveness,” they wrote.

“However, the AFP remains committed to ensuring children are safe on and offline and embraces any initiatives that work towards combating online child sexual exploitation,” the spokesperson said. “The AFP supports blocking as an initiative that protects the rights of the children depicted in images of abuse by preventing further distribution of these images. Removing access to images of exploitation and abuse is a valid tool in combating this crime. The AFP welcomes the voluntarily blocking scheme, in addition to the tools currently available to combat cybercrime.”

The AFP told iTNews that in some cases, due to the infrastructure being used, it may be cost-prohibitive or not technically possible or for ISPs to provide statistical data regarding their filtering. However, this week the AFP did not directly answer a question on whether it believed collecting data might not be technically possible. “Each participating ISP has advised if they can or cannot provide the statistical data, however the reasons may not have been provided to the AFP,” the organisation said.

In Senate Estimates hearings last week, AFP high-tech crime chief Neil Gaughan mentioned other ISPs had approached the AFP with interest in participating in the trial. However, this week, the AFP declined to detail which ISPs had done so, confirming that only the initial three ISPs had so far signed up.

“The AFP will continue to engage with them, as part of its overarching goal to disrupt the availability of child abuse material within Australia,” the spokesperson said. “As these ISPs are not yet confirmed, the AFP is not at liberty to mention which ISPs have approached the AFP with interest in participating. It is difficult to predict if and when these ISPs will participate in the trial. As it is a voluntary scheme, it is up to the ISP concerned to announce their participation in the trial.”

Telstra’s methodology in tracking web traffic blocks occurring during the trial has been questioned by independent commentators such as Michael Wyres; however the telco has not as of yet provided any further detail on how it collects its filtering data.

Delimiter has filed a Freedom of Information request with the AFP to request any email communication to or from Mr Gaughan over the past few months which refers to the Telstra results, with a view to obtaining further clarity about how the traffic was measured and how the results can be interpreted. The AFP has acknowledged the request and pledged to respond within two weeks.

opinion/analysis
If ISPs are not required to provide any statistics about how much traffic they block under their voluntary filtering schemes, it should be plain as day that there is simply not going to be any real way to measure how effective the voluntary filtering scheme is.

As I wrote last week, taken in isolation, Telstra’s statistics sound pretty onerous — 84,000 is a lot of requests for very disturbing material. However, we don’t know how these statistics were collected, or what percentage of Telstra’s overall traffic they represented. If it was something like one in a billion percent, for example, is that percentage worth enforcing a national filter system on every Australian, with the potential for scope creep which exists? We need to know more.

I am also disturbed at the lack of transparency which both Telstra and the AFP are demonstrating when it comes to this trial. Sure, right now the filtering trial only blocks sites which allegedly contain child pornography. But there is nothing to prevent scope creep to other subjects — such as borderline offensive but not illegal material — occurring; and the filter mechanisms are already in place.

If Australia is going to go down the filtering path, the nation needs more information about this entire process to decide whether it is a valid one. Right now all the discussions about it are being held behind closed doors, with no civilian oversight, and with no commitment to releasing detailed public information about it. The AFP’s disclosures this week in response to questions on the filter are a good start — but only that.

Image credit: Anja Ranneberg, royalty free

10 COMMENTS

  1. “Right now all the discussions about it are being held behind closed doors, with no civilian oversight….”

    Are they? Optus and Telstra have both announced what they are doing voluntarily. Customers are able to voluntarily change ISPs in response if they think it’s an inappropriate effort to restrict the distribution of such material.

    There is “civilian oversight” in that those that encounter a block page are able to appeal it. I’ve not heard any reports of anyone encountering the page and appealing. Have you?

    • Yeah, right.
      It is incredibly easy to unbundle oneself from a Telstra package with financial commitments for goodness knows how long, and it is incredibly easy to complain against blocked pages.
      Much easier than overriding the filter.
      [End sarcasm]

      The main thing annoying me with this affair is that 84,000 number that was plucked out of nowhere, literally. It is very obvious to anyone trying to think what that number represents that this number stands for nothing, at least nothing that has any meaning. Yet our politicians will gladly use the figure on the otherwise ignorant public…

      • Yep, appreciate that some customers may be in the midst of one or two year contracts. But if those customers are still keen, they can go to another ISP, at the end of their contract, that doesn’t want to restrict access to that sort of material.

        Are you speaking from experience on the appeal thing because the description in the article above does make it sound easy to challenge a block.

        I actually haven’t seen any reports of politicians using the figure – only Telstra.

  2. Plasma, invoking Godwin, with the best of intentions? lol.

    Telstra’s filter isn’t based on ACMA, it’s a list based on data from INTERPOL. ISPs don’t have to collect data because there’s no real policy framework that defines if they do or don’t.

    Telstra has decided to jump the gun, perhaps based on the notion that should mandatory filtering actually pass legislation, that they would (somehow) be immune. Bit like competition regulation.

    Also, being “disturbed” about the lack of disclosure comes with the territory.

  3. The interesting thing here is that the voluntary filter was supposed to be an IIA initiative commencing in July. http://www.iia.net.au/index.php/all-members/892-internet-industry-moves-on-blocking-child-pornography.html

    It is now being touted as an AFP trial. Does anyone really know what is happening? Why haven’t the other members of the IIA started filtering? Are there doubts that the Telecommunications Act section 313 authorises the filtering? Are we being fed a lot of male cow manure?

  4. “Sure, right now the filtering trial only blocks sites which allegedly contain child pornography. But there is nothing to prevent scope creep to other subjects…”

    I was surprised to see how the cleanfeed filter in the UK has been expanded into the area of policing copyright infringement – see http://www.bbc.co.uk/news/technology-14322957

    I assume this would be a blanket ban on what is actually just a search engine for usenet, rather than blocking specific links to infringing content. I wonder if this was ever intended to be in scope for cleanfeed when it was first touted?

  5. I posted similar concerns on another prominent web based IT site and was suggested by one responder of being ‘twitchy’ or ‘exposed’ and by another who obviously accepts everything that has been declared about the hidden ‘Interpol’ filter list being used as being unquestionable truth. They appear to be completely relaxed with the concept of collection of stats. Scope creep and hidden activity re this procedure seem to pose no issues for some. I posted two responses but unlike their responses mine are awaiting ‘assessment’ by a moderator, strangely for the first time as I recall. My issue there was how could their byline refer to “progress’ of the process being threatened by problems with stats collection when the voluntary filter process that was ‘quietly’ introduced by the three isps had no disclosure of this factor. maybe the latest comment by the AFP might get this a spect being discussed a bit more sceptically elsewhere.

    Like Conroys and other international ‘filters’ this thing gets progressively messier and more amateurish as the facts get exposed, but I guess the ‘anything is ok to save the children’ sheep and the ACL are pleased with the situation.

Comments are closed.