Letter On Fair Use

A letter written to a colleague working in the area of data privacy.

I would imagine that you're well aware of the Hargreaves Review .. but just in case

http://www.bbc.co.uk/news/technology-13429217
http://www.bbc.co.uk/news/technology-13422652

This was news to me and, from what I can see, its recommendations are sound. I agree with Rory Cellan-Jones' analysis that the winners are the content producers as opposed to content republishers, such as Google.

The doctrine of Fair Use is not a particularly good one. It's fundamental role in US law is making copyright law consistent with the freedom of speech assured by the First Amendment - otherwise copyright law could be struck down as unconstitutional. In other words, Fair Use is designed as a safeguard against prosecution for discussing a work.

Google have famously used the Fair Use doctrine, arguing that their colossal infrastructure for copying, indexing and republishing in summary form is protected as free speech in the same way as a reporter writing about a book or a song. The comparison appears to be ludicrous but the issue is how to make a robust distinction.

The Googlised interpretation of Fair Use is that no permission needs to be sought to (1) take a temporary copy of an entire work for automatic analysis (2) compile an index of an entire work and (3) republish an arbitrary number of summaries of sections of the work. The argument is that provided each individual summary would satisfy a Fair Use argument then this constitutes a legitimate analysis and summarisation process. And because this is legitimate to do once, it is equally legitimate to iterate this process billions of time.

The difference is that Google are leveraging the fact that they do not pay the face value of the books they index or the box office price of the films they review. They are leveraging an economic imbalance in the Internet, that published material is free because there is no micro-payment infrastructure.

The economy of the Internet is broken. A broadly similar situtation caused the rise of spam. Spam only arose because the sender only pays the cost of bandwidth (and spammers don't even pay that because they use botnets). When you receive a spam message, it costs the spammers nothing for you to receive email. It's free. If you had to put a stamp on unsolicited email, spam would never have been profitable.

The content of the Internet was created in the spirit of academic freedom. And there was a simple etiquette - you must freely return your knowledge in return for taking. The economics of the Internet reflect this unworldly innocence. When the world wide web happened, it wasn't simply the technophobes who were caught flat footed - it was everyone. Firms such as Google sprang up, taking 'free' content and republishing in order to make a profit.

As a thought experiment, let us suppose we had a micro-payment infrastructure on the Internet, so that every single web-site was behind a mini-paywall that could be configured by its owners. This is a tiny deviation from our existing model for consumers and a massive change for content republishers like Google. Making Google pay a fair cost for the content they index, as any non-Internet version of Google would have to, would significantly impact Google's business model.

Furthermore, the pernicious habit of caching website content to speed up delivery would be revealed to be the dubious practice it really is, it's a permission-free form of copying to improve an ISPs business value.

But this is only a thought experiment, so implementing world-wide micro-paywalls is not something we could ask for tomorrow. But excluding bulk-download of web sites for commercial caching or indexing purposes from the Fair Use provision would be a significant step forward in equalising the status of citizens and institutions.

My notion of Fair Use does not include automated copying, indexing and republishing entire works without permission and without compensation.