Emma Hart at Public Address has done an excellent post on how the Australian Government is spending $75 million implementing compulsory filtering of the Internet, and why we should be very wary of any such proposals here. Some extracts:
In July, the government ran a trial of various filtering systems in Tasmania. There’s an excellent round-up of the results here. In brief:
– while load testing was based on thirty users and only blocked 3930 sites, network degradation was as high as 75%. The more accurate the filter was, the worse the effect it had on performance. One filter caused a 22% degradation in speed when it wasn’t actually filtering.
– at best, sites were correctly blocked 92-95% of the time. At worst, more than one in ten got through.
– at best, sites were incorrectly blocked (blocked when they contained no objectionable content) 1% of the time. That doesn’t sound too bad, but imagine that’s your business, one of the one in a hundred sites blocked from the entire Australian market when you’ve done nothing wrong. At worst, over-blocking hit over 6%.
– the only way to filter content on instant messengers or peer to peer protocols was to block them completely.
– The filters do nothing to protect children from actual dangers such as cyber-bullying or stalking.
I helped set up a test a few years ago of the UK Clean Feed system, with the Dept of Internal Affairs and the Chief Censor’s Office. The UK system is actually quite good as they manually check every site that gets entered on – they don’t rely on some sort of “guess”. They didn’t appear to block any legitimate sites (a 1% false positive rate as the Ausises have is actually very high) but they didn’t block a huge proportion of objectionable (in the legal sense) sites. So relaying on such a filter may give a false sense of confidence, as the reality is that illegal sites changes hosts, domain names, IP addresses almost every day.
The way Australia is going about it, is just bureaucratic madness. Read Emma’s article in full to see why.