I’m really pleased to have had my review of the Court of Justice’s seminal UsedSoft v. Oracle decision published by the prestigious Society for Computers and Law. Click the image to go there.
I haven’t been able to write here for some time, but I have been writing elsewhere. Here’s a list of my recent copyright articles, with links. A list of my recent trade mark publications is here, and there’s a full list of all publications here.
YouTube does not cite the relevant DMCA language in its latest series of arguments because it has no answer to it, according to the reply letter brief filed yesterday in the Second Circuit by Viacom and The Football Association (et al.). In October, the parties in Viacom’s $1bn copyright lawsuit against YouTube for historic infringements of its audio-visual works were ordered by the Court to submit further arguments on: 1) whether “storage” under the DMCA includes syndicating files to third parties on commercial terms; and 2) how knowledge of “specific” URLs can be mandated in the DMCA’s “red flag” regime, and they have now completed their filings (here and here). Viacom is seeking reversal of a 2010 decision in which District Judge Louis L. Stanton, sitting in the District Court for the Southern District of New York, sent Viacom’s lawsuit to the cutting room floor without a trial.
The briefs make for interesting reading, and this was certainly Viacom’s most aggressive submission to date. There was plenty in both sides’ filings that I have not found persuasive, or have wholeheartedly disagreed with, but two points have clearly come out. Continue reading »
Fresh off Justia this morning is Viacom and The Premier League’s reply to YouTube’s court-ordered letter brief, as covered here. A copy of the filing in the Second Circuit is below and is well worth a read. Comments will follow later today.
Here’s YouTube’s post-argument brief on the extent of the “storage” exemption from liability under the DMCA, and clarification of how their assessment of knowledge under the statute can operate in a “red flag” regime. Comments will follow in due course, but an initial review suggests a slight rowing back from the position adopted at the hearing, but no real surprises.
File obtained via Justia.
Google has put an end to its controversial Keyword “Suggestion” Tool after the CJEU ruled that it might not benefit from the exemptions from liability found in the E-Commerce Directive. It also appears to have modified its remaining services to protect the “LOUIS VUITTON” trademark, but continues to allow advertisers to sully other luxury brands by offering obviously detrimental words such as “fake” and “replica”.
Google’s Keyword Suggestion Tool has been a source of controversy since its inception. In a nutshell, advertisers, regardless of their aims or track record, could use the Keyword Suggestion Tool to determine how popular other keywords were, so that they could add those keywords to ones already chosen for their advertising campaign. Where an advertiser had selected a generic term (such as “suitcases”), Google might suggest that they also use a trademark (such as “SAMSONITE”); where the advertiser had selected a trademark (such as “LOUIS VUITTON”) Google not only suggested obvious nouns such as “bag”, but also suggested words such as “imitation” and “copy”.
In defending Louis Vuitton’s suit against them in France, Google said it didn’t matter whether they were liable since their acts were automatically excused by Section 4 of Europe’s E-Commerce Directive, in particular Article 14 (the “hosting” safe harbor).
Court surprises Google by ruling that it may be held to account for helping advertisers to sell fakes after all
The French court asked the CJEU to clarify the scope of Article 14. The Court duly obliged, holding that only the “mere technical, automatic and passive” acts of an intermediate service provider would be excused by Article 14. This interpretation is not what Google had asked for, and is considerably narrower than the equivalent provisions of section 512 of the US Digital Millennium Copyright Act, which have been given a broad interpretation in Viacom v. YouTube (albeit only in the District Court so far). See further the discussion of Article 14 and Recital 42 here. Having given its interpretation of Article 14, the CJEU then instructed Paris that:
- “it is necessary to examine whether the role played by that service provider is neutral, in the sense that its conduct is merely technical, automatic and passive, pointing to a lack of knowledge or control of the data which it stores” (paragraph 114);
- “it is apparent … that, with the help of software which it has developed, Google processes the data entered by advertisers and the resulting display of the ads is made under conditions which Google controls. Thus, Google determines the order of display according to, inter alia, the remuneration paid by the advertisers” (paragraph 115);
- “the role played by Google in the drafting of the commercial message which accompanies the advertising link or in the establishment or selection of keywords is relevant” (paragraph 118).
Bing reveals Google’s name change. Changes to the program itself suggest that Google might have made its position worse
The quotes above, particularly from paragraph 118, appear to have precipitated Google’s subtle (but perhaps pointless) changes to its AdWords program. If you look at the following screen grabs, you will see that a Google search for “Google’s keyword suggestion tool” gives only one result from the “adwords.google.com” sub-domain. That page is titled “Keyword Tool – Google AdWords – Online Advertising by Google” (grab 17). But the result of the same search performed on Bing on today’s date gives the page a different title. Although the URL and the page description are both identical, the title on the page is “Google AdWords Keywords Suggestions” (grab 1). The conclusion must be that Google has changed the name of its program so recently that not all search engines have captured the change.
Although Google clearly felt that this change of name was necessary, this rose smells the same: Google is still offering the same keyword suggestion tool. While they may hope that dropping the (superfluous) acknowledgment of the program’s purpose from its name will persuade courts that it is not “controlling” the content of ads, other features of the revised program appear to have had the opposite effect:
- Note that Google reserves the right to “disapprove” any keywords added by advertisers – see, e.g., the “Important Note” at the bottom of grab 3. Either this is a bluff and Google has no intention of “disapproving” any keywords, or it is exercising control over the program in a manner that could take it outside Section 4 (see quotation from paragraph 114 above);
- Google also appears to be exercising control over which brands it will protect and which it will not:
- The owner of a website selling fake goods who selects the trademark “LOUIS VUITTON” as its keyword will not find any suggestions for “fake”, “imitation”, “copy” or any other damaging phrases (grab 3 and grab 4); BUT
- The owner of a website selling fake goods who selects the trademark “OMEGA” (grab 5) as its keyword, on the other hand, will be offered the keywords “replica” (grab 6, three times; grab 7, once; grab 8, twice; grab 9, five times;); and “fake” (grab 7, once; grab 8, twice; grab 9, twice), as well as offerings of competitors’ trademarks (grab 7 and grab 9). Choosing “ROLEX” (grab 10) as a keyword results in offers of “replica” (grab 10, four times; grab 11, five times; grab 12, six times; grab 13, eight times; grab 14, seven times; grab 15, six times; grab 16, five times); “fake” (grab 10, once; grab 12, once; grab 13, once; grab 14, once; grab 15, twice; grab 16, twice); and “imitation” (grab 11, once; grab 12, once; grab 13, once; grab 15, once; grab 16, once), as well as offerings of competitors’ trademarks (grab 14 and grab 15). This means that around half of Google’s “suggestions” for this particular luxury brand contain references to either fakes or competing products.
What is the takeaway?
Google appears to have acknowledged that its Keyword Suggestion Tool was a liability. Despite this, it seems to be reserving the right to control keywords as well as demonstrating that is willing to and does manipulate its keyword tool to control what can get through it. The fact that it has done this for one brand (“LOUIS VUITTON”) while failing to do so for others, even when the illegality of the suggested keywords is indisputable from the context provided by Google (see grab 18 and grab 19), must mean that Google’s keyword (suggestion) tool is not a merely technical, automatic and passive service. As a result, it should not be able to benefit from the Article 14 Hosting exemption. (Whether Article 14 even comes into play depends, of course, on whether the national courts find Google liable for something other than registered trademark infringement in the first place.)
Second, automated processes increase the frequency of action (and therefore infringements) beyond what a group of humans could ever do. It cannot be right that otherwise harmful acts can become permissible simply because they have been carried out by a program. Courts must remember that programs function according to the specific instructions they have been given, and that when a program acts it is only because someone has coded it to perform that precise function at that precise time – programs have no discretion; people do. Blanket immunity for computer programs invites programmers to ignore (or even turn a blind eye to) the potential and sometimes obvious damage their products can cause. “Neutral” services must therefore be interpreted to mean services which treat all inputs the same and whose output is not interfered with. Google’s actions in relation to the “LOUIS VUITTON” mark indicate either that the input is designated for special consideration or the results are vetted. Either way, the service is not “neutral”.
Third, advertisers risk purchasing a lawsuit from Google when they use the AdWords service. The CJEU has held in Google France, and reaffirmed in Portakabin, that an advertiser who selects a third party trademark as a keyword to trigger ads which are even vague as to their origin is likely to infringe those trademarks and be liable for damages, as well as to be the subject of injunctive relief and possibly a costs award. So unless an advertiser is able to set out clearly, within their limited characters, who they are and who is responsible for the ad and its related goods and/or services, and to avoid creating any perception that there is a link between them and the owner of the trademark, regardless of whether they reproduce the trademark in their ad or not, they should avoid using someone else’s trademark to advertise online.
After what seems like an eternity, the US Supreme Court today handed down its decision in Bilski vs. Kappos, on the patentability of various business methods. At first sight the decision does not seem to provide the overhaul of US patent law that many had been predicting. Much discussion will follow, so watch this space for additional commentary in the days and weeks ahead, particularly in relation to software patents (more properly “computer implemented inventions”).
The judgment (70-some pages) is available as a .pdf here. Enjoy!