ICRA - Censoring The World, Round 2

I've just skimmed through the recent releases of the "Internet Content Ratings Association", http://www.icra.org/

Frankly, I'm frightened in a way I haven't been for a while. This is the return of a rival approach for governments to censor the net, a system which I had thought was dead and buried.

Let me see if I can explain in a way that cuts through the pages of technical documents and the noise of press-release hype (I never put much credence in talk of parents - as a simple statment of fact, if it works for parents, it works for governments). Has anyone ever considered how difficult it is to censor the Internet? I have a saying - "It's a tough job to censor the world". But bright, smart, people are applying every bit of their technical skills to this daunting task.

There are two basic architectural approaches, which have competed for acceptance. The first one is the private blacklister. It works by having a business make up a huge blacklist of forbidden material, This is the current censorware system, euphemistically called "filtering". It's run by companies who distribute the blacklists, for a fee.

However, there is another approach, the "classification" system. In this scheme, a central organization applies various labels which indicate how freely the material will be allowed to be distributed - ranging from "available to everyone" to "banned". The MPAA ratings are an instance familiar to Americans, but that's not a simple example. While the MPAA ratings are technically mere private designations, they function as an industry-government collaboration (this is already a complex topic). Basically, "self-regulation" in place of government regulation (and backed by a fear of government action if the "self" regulation system is too lenient) is still a content regulation system - not movie reviews.

However, other countries, for example, Australia, don't fool around with privatizing the administration of the ratings system. There's the "Office of Film & Literature Classification" http://www.oflc.gov.au/ which runs the "classification" system.

This general model, based on ratings/classification, is favored by many European and similar governments. An earlier proposal from the UK can be read at "R3 Safety-Net" http://www.iwf.org.uk/about/rating_r3.htm (dates from September 1996).

Around 1997, there was notable competition between proponents of these two net-censorship approaches. To make a long story short, the censorware system (private blacklists) got to market first, with an implementation that functioned well enough to sell. While the "classification" proponents got mired down in legal and technical problems.

But some governments just aren't comfortable with the censorware system. They have large censorship, err, "classification", bureaus, and laws about ratings optimized to that system. The blacklist companies sometimes aren't very sensitive to the particular priorities of certain governments (i.e, France or Germany versus Nazi sites). So the European Union has been working on ratings-based censorship for a while. See http://www.europa.eu.int/information_society/programmes/iap/projects/filtering/index_en.htm

And they're back! That's the real news with ICRA. It's not just a ratings system. It's the second round of attempting to have a workable global ratings/classification-based censorship system.

You see, it's hard to censor the world, with the conflicting needs of several governments. There's a detailed mechanism to deal with the technical problem here. It's called PICSRules, and is a veritable "censorship cookbook". See http://www.w3.org/TR/REC-PICSRules

The examples there are just full of censorship goodies:

11 Reject any HTTP URLs from the www.badnews.com and www.worsenews.com hosts, and all URLs that specify a host whose ip address has 18 as its first eight bits (these are the addresses corresponding to mit.edu).

[Don't like 'badnews', it disappears! Or "sites outside the country"]

13 Specifies that documents which have an educational rating of 1 in the KP rating system (defined above) will be allowed. Documents which have no rating under this rating system, or which have a rating other than 1 will be examined according to the rules which follow.

[Now, replace that with "accepted news organizations" from the HomeOffice rating system will be allowed to pass]

14 Specifies that documents which have a violence rating of 3 or more in the KP rating system (defined above) will be blocked; explanatory text is provided for user-agents to display to users: after decoding, the text is: Blood's a "scary" thing. Documents which have no rating under this rating system, or which have a lower rating will be examined according to the rules which follow.

[Just imagine: "The government has determined this to be a terrorist site!"]

Note this PICSRules document is from 1997. Again, this approach was one which was put forth then, lost out, and is now being revamped and revived. See the references to PICSRules in the "Notes for software developers" in the "ICRAfilter" section of http://www.icra.org/

The cultural flaws of the ICRA ratings system are just a small part of the problem. The real danger is the attempt to get all the classification bureaus, blacklists, and whitelists, functioning in a censorship system. So that there can be a specification where, per the above "censorship cookbook", a mandate can effectively be put in place of "Ban these sites in general, but give a special exemption to this list of government-approved news organizations, but this other list of subversive sites is to be banned even if they claim to be news sites, etc."

That's been too costly for Western governments to implement. So far.

--
Seth Finkelstein
http://sethf.com
Seth Finkelstein's Infothought list - http://sethf.com/infothought/