Australia’s mandatory internet filtering by internet service providers (ISPs) won’t happen for at least two years. But we’re getting filtering anyway. Voluntarily. By ISPs. Next month.

Almost exactly a year ago, just before the 2010 federal election, communications minister Senator Stephen Conroy defused the controversy in a politically astute move. He responded to criticisms with process. He gave everyone a chance to feel like they’d achieved a minor victory, even though the underlying policy hadn’t changed.

Concerned that the Refused Classification (RC) content category wasn’t really just the worst-of-the-worst? Let’s review that. That’ll take a year or two. In fact the Australian Law Reform Commission (ALRC) is now reviewing the entire content classification system across all media.

Concerned that the Australian Communications and Media Authority (ACMA) isn’t so flash at classifying content and compiling the blacklist of prohibited online content? Let’s pay attention to the outcome of consultations on transparency and accountability for ISP filtering, create a new framework for assessing complaints from the public, and get the Classification Board to do the classification. That’ll take a bunch more time once the ALRC’c review is completed.

Concerned that there’s still child exploitation material on the web? Let’s make a start on that. Let’s get ACMA to pull the kiddie-p-rn from their full blacklist, and see if ISPs will at least block that voluntarily. They could start work straight away.

Telstra, Optus and Primus volunteered, representing about 70% of the internet-using population at the time. It’ll take a year, they reckoned. And that brings us to today.

Telstra and Optus are expected to have their filters ready within weeks, although the situation with Primus is unclear.

But instead of using ACMA’s blacklist, Telstra will use a list compiled by Interpol. This is presumably the list announced by Interpol in October 2010. Designed specifically for use by ISPs, it lists internet domains containing “severe child s-xual abuse content”. According to pan-European Cospol Internet Related Child Abusive Material Project (CIRCAMP) guidelines, children depicted must be real — computer-generated or drawn images are not included, so Simpsons s-xual parodies don’t count — they must be or appear to be under 13 years old, and the material must depict contact of a s-xual nature.

In other words, the Interpol list really is the worst of the worst, and nothing else.

The Internet Industry Association (IIA) is also about to release a voluntary industry code that would see an estimated 80% to 90% of Australian internet connections filtered by the Interpol blacklist over the next year. Attempts to access domains on the list would be redirected to an Interpol block page.

“ISPs who block access to sites would be doing so in accordance with a legal request for assistance under Australia’s existing Telecommunications Act (section 313); no new laws will be required to implement this scheme,” the IIA said in a statement. “Because of the possibility of accidental access to blocked sites, users will not be tracked or reported under the scheme.”

Such blocking has virtually no effect on internet speeds. Every website visit requires a look-up to translate the domain example.com into a numerical internet (IP) address so the connection can be made. If the domain is a blocked site, the look-up merely returns the address of the Interpol block page instead of the site’s real address.

Even ACMA’s blacklist, with individual web addresses (URLs) rather than domains, can be filtered with techniques that are significantly faster than those tested three years ago. Blocking is a two-stage process. First, all traffic passes through a router — the devices that direct traffic around the internet. Connections to an IP address that hosts any blacklisted content is shunted sideways to a second stage of processing. Only traffic to those address goes through the deeper inspection of which URL was requested, letting through everything but the blacklisted ones.

Once the IIA’s voluntary filter is in place, the government’s mandatory filter may never happen. As John Hilvert reported at iTnews, the Joint Select Committee on Cyber-Safety declined to make any recommendations or endorse the policy.

“There is no evidence of reluctance by ISPs to take down Refused Classification material,” the committee wrote. Large web hosting companies already take down a much wider range of “inappropriate” content. And ACMA surveys show that between 40 and 50% of parents already use internet filters at home, again covering a lot more than RC.