Scraping and the Craigslist case

A series of recent Australian cases have exposed the potential inadequacy of copyright laws to protect databases and other compilations of factual information from wholesale copying. This inadequacy is exacerbated online, where content can be scraped and reproduced easily and quickly by third parties using automated software. Whilst legislators in some foreign jurisdictions have enacted specific laws to protect investment in databases (such as the sui generis database right in Europe), Australia has been slow to react and technological advances seem to outpace the law.

As a consequence, proprietors of free access “classified” websites increasingly rely on alternative methods to protect their information assets. This includes reliance on website “Terms of Use” (“TOU”), usually accessible via a hyperlink at the bottom of a webpage, which purport to form binding contracts between the website operator and website users (including scrapers). We are waiting to see how Australian Courts will enforce these terms and conditions in contract law.

Ongoing litigation in the United States involving global classified’s heavyweight Craigslist shows that Australia is not alone in grappling with the legal status of website scraping. The Craigslist case provides useful insight into the types of claims that might be made against alleged scrapers if proceedings were issued in Australia.

The Craigslist case

Craigslist is one of the world’s largest classified websites. According to the ‘About Us’ section of its website, Craigslist has more than 50 billion page views per month.

Craigslist monetises its website by charging fees for job postings and apartment listings, but only in certain strategic locations. For other classified advertisers, the site remains free.

As with most free to use online resources, website traffic is the lifeblood of Craigslist. It is the website’s reach that Craigslist markets to prospective advertisers. Therefore, it is not surprising that Craigslist has commenced proceedings in the US District Court of California against three companies, all of whom allegedly scrape the Craigslist website in an attempt to divert website traffic away from Craigslist to its competitors. The Defendants in this case are:

3taps Inc, which allegedly scrapes all advertisements from the Craigslist website in real time, in order to provide a “Craigslist API” to third parties, (enabling the third party to incorporate advertisements from Craigslist in their own offerings, such as mobile apps). 3taps also operated a website at the domain name craiggers.com, which Craigslist alleges “essentially replicated the entire Craigslist website”;
Paddmapper Inc, which operates a property portal website advertising real estate listings for sale and rent, and largely consists of listings originally posted to Craigslist. Paddmapper is alleged to have scraped Craigslist and then republished the listings on its website; and
Discover Home Network Inc (trading as Lovely), which operates another property portal that includes listings originally posted to Craigslist that it receives from 3taps via the “Craigslist API”.

One of the obvious commercial effects of this conduct is to reduce the size of Craigslist’s marketable website traffic statistics.

Craigslist’s claims against the three alleged scrapers include:

breach of contract (for breach of the Craigslist’s TOU, which prevent scraping);
trespass to goods (for unauthorised interference with Craigslist’s computer system, as large scale scraping causes the computer system to slow down);
copyright infringement (in both the user generated posts themselves and in the compilation of those posts, notwithstanding obvious issues in relation to subsistence and ownership);
trade mark infringement (for using the registered mark CRAIGSLIST as a badge of origin);
passing off, dilution of a famous mark, and unfair competition (all related to the third parties’ use of the CRAIGSLIST mark); and
violations of the Computer Fraud and Abuse Act (“CFAA”) (for intentionally accessing a computer system without authorisation).

In response, the Defendants applied to the Californian District Court for summary dismissal of a number of these claims. This required the Court to consider whether Craigslist had at least a “plausible” case in relation to the pleaded causes of action.

Interestingly, the Defendants’ motion to dismiss was, for the most part, denied by the Court in April this year. The Court thought that Craigslist had at least a plausible claim in relation to (amongst other things), breach of contract, trespass, copyright infringement, trade mark infringement and violation of the CFAA.

Below are some of our observations about the more interesting aspects of the interlocutory decision, and issues that an Australian Court might have to grapple with.

Breach of contract

Craigslist’s TOU prohibit scraping.
Users of Craigslist who post advertisements are required to accept the TOU before posting (i.e. by clicking on a box), however, it is unclear whether there is any similar mechanism for other users (including the Defendants in this case) to accept and agree to be bound by the TOU.
Notwithstanding this, the Defendants did not argue in their summary dismissal application that the TOU did not form a binding contract with them, i.e. through lack of positive acceptance. Rather, they (unsuccessfully) argued that the contract claim was invalid because of a provision in the United States’ Copyright Act that is designed to ensure that there is no inconsistency between State and Federal intellectual property laws. No similar provision applies in Australia.
An Australian Court will, at some point, need to consider the legal effect of website TOUs.

Until then, we pose this: does the apparent acceptance of the contractual force of TOUs by the scrapers in the Craigslist case represent a shifting tide?

Trespass to goods

The United States has for some time recognised that it can be a “trespass to goods” if one interferes with the possession of a computer system (a “chattel”, which hosts a website) in an unauthorised manner (e.g. unauthorised scraping) where this causes damage to the computer system (e.g. by slowing it down).
Here, the Court had to determine whether Craigslist had a plausible claim of actual damage resulting from the Defendants’ scraping.
Given that 3taps boasted about its ability to copy tens of millions of Craigslist postings in real time, the Court found that “it is plausible that such access could divert sufficient computing and communications resources to impair the website’s and server’s functionality” and so rejected the Defendants’ claim for summary dismissal of Craigslist’s trespass claim.
Theoretically, there is no reason why trespass to goods law would not also extend to cyberspace in Australia. In Australia, the tort has not traditionally required any damage to the good in order to be made out. Australian Courts might seek to confine the doctrine of trespass on the internet in some way, for example, by requiring some damage to the value of the relevant computer system or website to be shown (ie. by significantly slowing it down).

Copyright

The content of the Craigslist website was pleaded as falling into two categories of literary works:

the user-generated listings and posts (eg: “two jousting sticks for sale – barely used!”); and
the compilation of those posts, as arranged by Craigslist.

Again, it appears that the Defendants accepted that copyright subsisted in the user-generated posts themselves. The question for the Court was whether Craigslist, not being the author of the posts, had the right to sue if copyright in them was infringed.
The Court’s preliminary finding was that, without anything in the TOU granting Craigslist exclusive rights in relation to the user-generated posts, Craigslist had no basis on which to sue third parties for copyright infringement. However, there was a brief period during which the TOU did grant an exclusive licence to Craigslist, and so the Court found that Craigslist had a plausible copyright claim for that period.
In relation to the compilation, the central question was whether the work was sufficiently original for copyright to subsist. Making a preliminary finding in Craigslist’s favour, the Court stated that “Craigslist, in deciding which categories to include, and under what name, displayed some minimal level of creativity” and therefore that Craigslist had at least a plausible claim to sue for copyright infringement.
The threshold for originality cited by the US District Court in this case appears lower than in Australia, where it has been held that an author of a compilation must exercise “independent intellectual effort” of a literary nature in order for copyright to subsist. [Ed: We found this part of the decision quite surprising although note there will be full argument on the question in due course.]

As this part of the Craigslist case was a motion to dismiss, rather than a final hearing, it does not provide any determinative views about whether the Defendants actually infringed Craigslist rights by scraping and reproducing classified listings. Rather, it merely indicates that Craigslist’s claims are “plausible” under US law.

A trial in the Craigslist case is likely to take place later this year, with publishers of classified websites around the world awaiting judgment with bated breath. Stay tuned.