According to ABC News, The Register, and a few others. On Monday Labor let the cat out of the bag, wait no there is no cat, and certainly no bag in which a cat has been placed. It seems that in its enthusiasm of the new government is pushing through “mandatory internet filters to protect children”. While there is still a lot of obvious ground work to be covered and no legislation drafted as yet, it looks like the ball is rolling.

Naturally there are concerns:

  • Privacy: Big Brother filtering could mean Big Brother logging.
  • Speed: Filtering requires at least one extra step processing HTTP requests.
  • Freedom of Speech*: The “clean feed” is opt-out. Why do we have to opt-out to get freedom?

* I’m not sure that freedom of speech is really at risk. After all for the filter to impose that way, sites that are not specifically explicit would need to find their way onto the black list. While that is not impossible, there would be severe democratic consequences.

Do I think this is a good idea? The intention is good, but the action is focused in the wrong area. So, No.

Technologically speaking, black lists don’t work – but there is no way the system could do real time filtering: NLP is very complex and there is no way for a computer program to “look at” content and decide it’s appropriateness – therefore we are stuck with black lists. Also a server side setup is relatively easy to bypass: use ssh to forward all traffic to a proxy outside the filter zone and you’re free to do what ever you want. It took less than 30 minutes to break the multi-million dollar filter last year, if we setup black lists, the bypass already exists!!!

Don’t get me wrong: I do believe that aim of the filters is a very worthwhile cause. It’s would be my expectation that, if the government is serious, they should spend time and money on producing/pioneering a solution that works. They have asked “how can we use black lists?”. I suggest taking a step back and ask how can we do this?