Remedies for Web Scraping
North American courts are not really sure what to do about web scraping, where someone uses automation to take information in bulk from a web site and puts it on his/her/its own site in competition with the original site. Sometimes (but not often) doing this is held to be trespass to chattels. More often it has been held to violate the terms of use of the victim site, even if those terms are ‘webwrap’, i.e. never explicitly consented to. (In fact, it is freakishly rare for webwrap conditions to be upheld except against scrapers).
The federal court in Illinois recently had to deal with such a situation. A dog lover put together a site with information about dog pedigrees for a certain breed. Someone else scraped that data, along with much other data, and assembled a web site with similar information. The original site owner put a notice on her site urging people to use her site as it added more value than the other, and referred to the owner of the second site as a thief who had stolen her data.
The owner of the second site sued for defamation, on the basis that information could not be stolen, so he was not a thief. No doubt the lack of a ready civil remedy reinforced his case. He said that the original owner’s lawyers had advised her that she has no remedy.
However, the judge was prepared to call a spade a spade, and a thief a thief, or at least to allow the original owner to do so. The fact that the law might not provide much of a remedy did not mean that the term was inappropriate. In any event the term ‘thief’ was not used in a technical sense but in a common understanding as someone who had taken something he was not entitled to. The original owner also had a qualified privilege in her remarks. It was not demonstrated that she published her remarks with ‘malice’, i.e. knowing or believing that they were false. She was entitled to disagree with her lawyers.
Would the Canadian law of defamation produce the same result? Am I right in thinking that Canadian IP, contract and criminal law do not differ substantially from US law on the relevant points, i.e. it would be hard for the original site owner to do much at law if someone scraped the content? The only Canadian cases I know of on point are both about scraping real estate listings, and both turned on what to me is the uncomfortable ground of a webwrap ‘agreement’ not to scrape…
Would the French law of ‘parasitism’ be a useful addition to Canadian law? It seems to allow judges to punish any commercial behaviour they don’t think is fair, regardless of intellectual property rules or other law.
The Canadian decision you are talking about is here.
The idea of judges deciding what is fair commercial behaviour is a bit unsettling – it would have the effect of stifling free enterprise in Canada. It is clear there is no blanket ownership right over information in Canada. If the information is not protected by copyright then I don’t see why a court should set in to prohibit scraping. Copyright (as with all IP rights) involves a balancing act between promoting innovation and encouraging competition and for a court to restrict copying based on unfair commercial behaviour where the legislature has otherwise said should be permitted is to upset that balance.
In any case, the concept of “fair dealing” in Canada, as the name suggests, already incorporates notions of fairness in the copyright context. The whole purpose of the fair delaing test is to determine whether the permitted dealing is fair. So the courts can and do get to evaluate the fairness when deciding whether or not to grant an exception to copyright infringement.
The other case is out of Quebec – Sutton Realty.
I am not sure that ‘fair dealing’ covers the kind of sleazy behaviour that web scraping often involves – certainly the use made of the scraped information does not fall within what has been typically held to be fair dealing. It’s out-and-out commercial competition, usually – but from a data base so maybe not so easy to deal with under copyright law.
I find the French decisions on parasitism pretty paternalistic, and would not really recommend adopting that principle here – but at least it is a remedy based on what is wrong with the practice, rather than inventing a contract without notice or consideration then enforcing it.
We have seen a huge crackdown from a legal perspective in the US on how the courts view web scraping. I outlined the trend in cases in a recent blog post that I think is relevant: http://www.distilnetworks.com/is-web-scraping-illegal-depends-on-what-the-meaning-of-the-word-is-is/