Rights for Robots: Repurposing the First Amendment

Introduction

We spoke in class of the two types of people: those who know how to change the behavior of computers, and those whose behavior computers change. Here I wish to examine one way in which the laws that were meant to protect the latter can be abused to further entrench the former. As technology advances faster than the law, it is important to take note of how new behaviors might be classified incorrectly within our existing legal regimes. A good example lies in an area of First Amendment jurisprudence. If we believe that it is as important to prevent the First Amendment’s misapplication as it is its erosion, then this will be an issue worthy of our attention.

Background: The Search King Case

In 2003, in a case called Search King, Inc. v. Google Technology, Inc., Google was sued for actively demoting a website’s Page Rank. Google quickly responded that its search results constituted editorial opinions, and were afforded full free speech protection under the First Amendment. Search King, on the other hand, argued that the Page Rank method is a patented, mechanical system incapable of rendering subjective opinions. Since ideas are not patentable, and since patented processes must be replicable, the Page Rank system must be entirely objective. The court sided with Google, granting the company’s motion to dismiss the claims against it. While the facts of this particular case did necessitate such an outcome (Google committed no legal wrong by manually adjusting Search King’s rank), the court took its analysis one step further, holding, quite unnecessarily, that “Google’s Page Ranks are entitled to ‘full constitutional protection.’” The court should have avoided the First Amendment analysis altogether under the cannon of constitutional avoidance, but the decision it did make on the issue represents a horrible perversion of the Amendment and its purpose.

First Amendment Protection for Automated Search Results

Google has since reiterated its stance on the First Amendment issue as it faces a new era of potential regulations and interventions. In anticipation of the debate, Google commissioned legal scholar Eugene Volokh to compose a White Paper arguing for free speech protection of search rankings. Volokh relied in part on the Search King decision, characterizing search rankings as editorial judgments similar to those reflected on the front page of a newspaper.

Doctrinally, this approach is both inappropriate and damaging. While Google’s users enjoy First Amendment protection for content found by search, and while Google enjoys protection for the original content that it creates, its algorithmic search rankings (that is, the specific order in which results appear) are not protectable speech. Academics have offered helpful explanations. One approach argues that, while a search ranking may seem like an opinion, it is no more protectable than the opinion a car alarm expresses when it senses danger and elicits the corresponding output of loud noises. While a newspaper’s mission is to communicate ideas, a search engine’s purpose is to locate information in a database. In this sense, the search engine is more like a telephone switch connecting users than a newspaper—it does not produce speech so much as functional communication. Another approach suggests that Google search is best thought of as an index of the web, most similar to card catalog schemes found in many libraries, which use the Dewey Decimal System. (Courts have in fact relied on the card catalog analogy in the past to understand meta tags).

Why This Matters

No matter how one understands it, the policy implications of allowing Google to inappropriately hide behind the First Amendment are startling. In 2003 in Oklahoma district court the issue was Search King, and it went largely unnoticed by the general public. But in 2013, with the world watching, the issues range from antitrust to privacy protection and have a great effect on public policy matters. The Federal Trade Commission recently completed an investigation that examined the core of Google’s business practices. The rank manipulation problem has become much more important as questions have focused on the extent to which Google lists results for its own products more prominently than competitors’ pages.

The F.T.C. ultimately settled with Google on a few small issues and ended its investigation, a decision that surprised many observers. The commission could not prove that Google “changes its search algorithm to purposely harm competitors and favor itself,” which antitrust law requires. Nevertheless, this is not the end for Google—many expect that the F.T.C. could return to the matter in the future, as they have done in past instances. If that is the case, we can be sure that the First Amendment defense will remain a formidable weapon in Google’s back pocket. If Google’s automated results are protected opinions, it would give Google an impenetrable shield against attempts to reign in its power and protect consumers.

The Future

Antitrust is only one area that could be affected. Inevitably other consumer protection agencies will examine Google’s practices, and privacy issues will emerge relating to information and data gathering. As we rely on computers to make more and more decisions for us, the question of free speech for automated processes will become crucial, not just for companies like Google, Facebook, and Verizon (all of whom have relied on First Amendment defenses to both common law and regulatory claims) but also for consumers and citizens. We need a strong and robust First Amendment aimed at limiting power; we do not want those in control to repurpose the Amendment to perpetuate power.

This argument makes no sense to me. Is Google entitled to say that it is privileged to avoid a defamation or tortious interference with business reputation claim because it has expressed an opinion? Apparently you agree that it is, as I certainly do. To the extent that search results are opinion, they are privileged against tort claims under state law that are "like" defamation in alleging reputational harm. This is surely a correct result, which depends not on whether Google formed the opinion using only a computer or also one or more human brains. The entity being sued is Google, the opinion is Google's opinion however they came by it, and that's the end of the story. See, for a useful example, Trump v. Chicago Tribune Co., 616 FSupp 1434 (1985).

That doesn't mean that they are also shielded from every other kind of liability. Google's opinions aren't therefore shielded from antitrust liability, if their expression or implementation somehow involves the unreasonable restriction of interstate commerce, or from liability for unfair competitive conduct under the Federal Trade Act: an "opinion" can obviously be false advertising, for example. Google's "opinions" are fully regulable under the security laws, too, for example, if they somehow fall within the regulatory ambit established by the relevant statutes. To say that their opinions are entitled to full First Amendment protection is not to deny any of those propositions. We know, to take another example, where one can express an opinion, however one reached it, with immunity from antitrust liability for expressing it: to the legislature. This, the infamous Noerr-Pennington doctrine, does not require one's lobbyist to be a human being, or one's research to be performed without machine intervention, nor should it. The existence of this rule has no effect on preventing "opinion" from being part of a collusive conspiracy to set prices, on the other hand.

So I don't understand what's going on. One District Judge said one thing that could under some other circumstances be capable of confusing generalization, in the middle of an opinion everyone agrees is right. Of course Google paid Volokh to emit supportive-sounding crap that could only help and not hurt them in other situations. Of course Volokh provided the crap: it doesn't matter to him whether he's talking sense or not, so long as he's getting paid and can listen to the sound of his own voice richly amplified by Google's riches. But nothing they have said through one another's mouths is wrong, or important. Google wants to have its right to rank search results as it pleases constitutionalized wherever possible. That's fine, and will have no technical effect on the other issues you discuss, whether in the US or abroad, anytime, except insofar as Google is entitled to have its communications judged by the same constitutional standards in the US courts as everyone else. What's the big deal?

• #Set ALLOWTOPICVIEW = TWikiAdminGroup, AndrewReich
• #Set DENYTOPICVIEW = TWikiGuest