In its Equustek ruling in June, the Canadian Supreme Court held that Google must delete search results for users everywhere in the world, based on Canadian law. Google has now filed suit in the US, asking the court to confirm that the order can’t be enforced there. Here’s my take on that claim.
It’s important to grasp a few things that aren’t happening. For one thing, this isn’t Google thumbing its nose at Canadian courts — though I worry some may perceive it that way. And no one is saying that US courts can tell Canada what Canadian law is. Finally, it isn’t primarily a question about whether US courts will enforce the Canadian judgment. Plaintiffs haven’t asked US courts to do that, and I doubt they’re likely to. Canadian courts can enforce the judgment just fine – by seizing Google’s Canadian assets, shutting down its market access, or even arresting employees. In its ruling, though, the Canadian Supreme Court effectively invited Google to come back and show lower courts that the law diverged in other countries, particularly with respect to free expression. So the move is in part an effort to do just that.
I see this as one part policy play, one part US legal play, and one part Canadian procedural play. The policy play is simple: the suit calls more attention to the alarming precedent of one country imposing global limits on people’s access to information. That’s important in this case, and even more important in the pending French “Right to Be Forgotten” case seeking global de-listing before the Court of Justice of the EU. More state institutions, including legislatures and ministries of state, trade, and justice in countries around the world, should be part of this conversation. So far, they’ve left the hard work to the courts, without giving them adequate doctrinal tools.
The legal parts are more complicated.
The substantive US legal arguments are based on international comity; the First Amendment; the Communications Decency Act Section 230 (CDA 230), which immunizes intermediaries for most removal demands; and Preliminary Injunction standards. They are interesting claims, and not ones I think Google would go out on a limb for absent this case.
The CDA 230 claim will force the court to look at questions about the statute’s intellectual property carve-out that might, from the perspective of Internet companies, be better left unexamined. The Ninth Circuit has said that CDA 230 exposes intermediaries to liability only for US federal IP claims. Its reasoning has been widely criticized, including by other courts, which have said intermediaries can be liable for state IP claims as well. Asking the court about Canadian IP claims and their US analogs in this case will stir up this debate again.
Google’s First Amendment claim has a straightforward component based on a 1999 case, Ford v. Lane. The court there rejected a trade-secret-based injunction on First Amendment grounds. More intriguingly, Google’s argument builds on a line of cases confirming search engines’ own First Amendment rights to select search results. Those cases were all brought by plaintiffs upset about being excluded from search results, though. Google’s argument extends the theory, claiming a right to include search results that, according to plaintiffs, violate the law. Personally, this version of the First Amendment argument makes me a little uneasy. Many people think platforms can’t be active speakers for US First Amendment purposes but still claim immunity as “passive conduits” under Intermediary Liability standards in Europe and elsewhere. As a logical matter, both should be possible: Google can express editorial judgment in its selection of results without adopting the content of any particular page as its own speech. (By analogy, an editor’s selection of short stories for a compilation is copyrightable expression, distinct from the authors’ separate copyrights in the stories.) But that logical distinction hasn’t gotten much traction in Europe. And it’s simpler to make when Google excludes search results, since then there is no issue of liability for webpage content.
As a matter of civil procedure, I think Google is right to say that a Preliminary Injunction like this would never have issued from a US court. Plaintiffs didn’t establish any rights enforceable outside of Canada, for one thing. Canadian intellectual property experts have suggested that, under applicable law, no such rights exist. (Inattention to the merits of plaintiffs’ claims is widespread – many analysts seemingly take for granted that plaintiff’s claims should be universally valid, typically based on loose analogies to copyright.) Plaintiffs also didn’t explain why they shouldn’t just pursue alternate channels for more effective relief, like asking web hosts to stop hosting the allegedly infringing sites. And, importantly, the court did not engage in the kind of public interest balancing that would have been required in the US.
So, will Google win? It’s not an easy question. They might fail on procedural grounds in the US, and not even get to the merits. This case reminds everyone of the Yahoo France case at the turn of the millennium, which ultimately died a procedural death. But there are some important differences. In the Yahoo case, a French court ordered Yahoo to stop showing Nazi memorabilia auction listings to French users. Like Google, Yahoo sought declaratory judgment in the US. A federal district court initially found a First Amendment violation. But after a couple of appellate rounds, the 9th Circuit sitting en banc dismissed the case. The opinion was messy, with no majority for any one analysis – three judges wanted to dismiss for lack of personal jurisdiction, and three would have dismissed on ripeness grounds. That was enough to make the case go away with no substantive resolution.
Google’s case looks like Yahoo’s inasmuch as the company is saying “the foreign court is forcing us to have US employees reconfigure our product on US servers.” But the important difference is that, unlike in the Yahoo case, Google’s foreign order will affect American Internet users – suppressing information they would otherwise have found in search results. That’s a big deal. And for what it’s worth, in the Yahoo case, French plaintiffs disclaimed any intention of enforcing the order here, which mattered to the US court. Overall, I don’t think the Yahoo case tells us what to expect now.
There’s also the question of what winning would mean, which ties closely to the Canadian procedural issues. Google is asking for declaratory judgment and for an injunction against enforcement of the Canadian order. It’s not clear what it would mean for the US court to “enjoin enforcement” of the order. If it meant actually ordering Google not to comply, that would create a hard conflict of law to show the Canadian courts. But an order to keep these search results intact would raise other problems – including, per Google’s arguments, violating the First Amendment. Perhaps there is some needle-threading to be done, with an injunction against removing results “on the basis of the Canadian order.” That would create a hard conflict between national legal obligations, but it would be weird. Alternately, an injunction could just say that plaintiffs can’t seek enforcement here. But it doesn’t seem like you should need an injunction for that – and it wouldn’t create a hard conflict with the Canadian order.
If Google does succeed, it can go back to Canada and say, “see, we told you the law in other countries was not consistent.” That would bring a couple of deeper issues into focus in Canada.
First is the question whether Canada would limit its order based on a “soft” conflict (a determination that US courts would not have ordered removal) or only based on a “hard” conflict (a determination that Google would violate US law by removing). The hard conflict standard is the one I remember from international law. But that standard is a problem when speech and information rights are at issue, because those rights will always lose out. Constitutional free expression guarantees in the US and elsewhere don’t prohibit private parties from suppressing speech, so there is no hard conflict. If courts issue extraterritorial removal orders in this situation, then Country A’s speech-suppressive laws will always trump Country B’s speech-protective laws. That can’t be the right outcome. (I think Dan Svantesson has written on this topic, by the way.)
The second interesting procedural issue, if Google comes back with evidence of conflicting US law, is the absurdity of giving Google the burden to make this showing. It seems obvious that plaintiffs should bear this burden. They’re the ones asking the court to reach across borders and exercise its power. It doesn’t seem right to make a defendant go research the world’s laws, and it seems even less right for an “innocent non-party,” which is Google’s role in this case, to do it. (And saying “Google should do it because it’s rich” is a cop-out, absent some systematic, economic-justice-based overhaul of civil procedure.) I’m not sure if there is room left, procedurally, for the Canadian courts to revisit this question. But maybe at least other national courts considering following in Canada’s footsteps will be given pause.
It’s a good time to be an Internet jurisdiction nerd. (If you are one, check out the great panels we had at our Stanford conference on the topic last year.) It’s not a great time to be an Internet user trying to speak or access information. The Canadian ruling may have opened the floodgates for more countries to impose their own values and speech laws – from Saudi blasphemy laws to French privacy laws to Russian anti-gay laws – on the rest of the world. The new US case is an important last-ditch effort to stave that off.
Disclosure: As Associate General Counsel at Google, I was involved in setting and defending the company’s removal policies at issue in this case, and worked on it at the Court of Appeals stage.
This post originally appeared on the Center for Internet and Society blog and is reproduced with permission and thanks.