Are Google’s Algorithms a Form of “Free Speech”?

The ongoing debate as to whether Google, Inc. should be considered an information provider or ISP as defined under Section 230 of the Communications Decency Act or a publisher akin to media entities such as The New York Times is becoming increasingly more vigorous. With a growing number of conservative and alt-right voices being censored, the U.S. government has resurrected its threats of regulatory intervention against interactive sites such as Alphabet, Inc., the parent company of Google, arguing that a lack of competition is affording Google and others unfettered power to dictate speech.

But even if the Federal Trade Commission and Department of Justice determine that Google is in fact violating antitrust laws and/or the company’s Section 230 status is challenged, a fundamental question would still persist: Are Google’s search engines a form of free speech and therefore entitled to First Amendment protection, notwithstanding any limitation set forth under Section 230?

The Federal Trade Commission’s 2011 investigation into the company’s alleged anticompetitive activities within the vertical market of digital advertising prompted an examination of whether Google’s practices constitute free expression. Renowned First Amendment scholar and UCLA professor Eugene Volokh argued that information provider services such as Google qualify for the same First Amendment protection as an individual or the press.

Referencing extensive case law, Volokh’s White Paper (commissioned by Google) entitled “First Amendment Protection for Search Engine Search Results” postulated that as a search engine, Google, via its employees, exercises “editorial judgment” in order to collate search results and design computer algorithms, similar to how a newspaper determines what news to feature and how to assemble the same. And this “judgment” is constitutionally protected, including any bias it may demonstrate.

Google’s “speech,” therefore, is expressed in the form of its search results. And the methods Google utilizes to prioritize such speech and whether it decides to include certain results are subject to First Amendment laws.

Simply put, Google, as a privately-owned corporate entity, has the same right to exercise its free speech in the form of search results as an American has the right to selectively express their thoughts using words as well as the freedom to remain silent. The Supreme Court recognizes that Internet speech is entitled to the same constitutional protections as those afforded to individual speech or newspapers.

Specifically, the First Amendment includes how material is selected and arranged, whether fact- or opinion-based. It also indemnifies computerized algorithms that yield these search results, even those manually manipulated by humans, because the Constitution protects a speaker’s right to include and exclude content, even if such content is considered unfair to others. (See Hurley v. Irish-Am Gay, Lesbian & Bisexual Group of Boston, 515 U.S. 557, 573 (1995)) and Miami Herald Publishing Co. v. Tornillo, 418 U.S. 241, 258 (1974)).

Although Eugene Volokh was hired by Google to offer his academic opinion, his findings, subjectivity notwithstanding, are tethered to extensive case law and legal precedent that remains meritorious 7 years later.

In 2011, Google was the subject of a FTC probe into whether the company intentionally used its behemoth market presence to promote its products over those of its competitors. Top competitors, namely, Microsoft, and legal scholars argued that Google used its market dominance to unfairly prioritize its own proprietary products and services over those of its competitors.

The FTC concluded its investigation in 2013. Although the Commission did not find Google had violated antitrust laws, the company agreed to change its business practices to assuage the FTC’s concerns about competition. In short, the agreement provided that Google would give online advertisers greater management over its Google’s AdWords and rival ad platforms by removing restrictions which precluded advertisers from managing their ads.

Since the publication of Volokh’s White Paper in 2012, Google has become far more than a search engine platform. In similar company as social media sites Facebook, Inc. and Twitter, this collective monolith uses its status and market dominance to not only provide access to information but also to suppress arbitrary “hate” speech. These companies all but require users to give away their rights to free expression under oft changing terms of service in exchange for access. Consequently, many allege that executives of Google and other interactive spaces have appointed themselves as arbiters of “hateful” content.

The solution offered by those who oppose government intervention is to use alternate services. This suggestion is inherently flawed based on the fact that there exist few, if any, viable alternatives. Such a simplistic view denies the degree of influence Google and others have not just as service providers but also as collectors of user information. These entities have amassed a tremendous amount of personal data that in turn generates billions of dollars in digital ad sales annually.

For example, in recent months, YouTube, which is owned by Google’s parent company, Alphabet, Inc., has purged and/or threatened to purge a number of conservative-based channels (e.g., Alex Jones, InfoWars, and Steven Crowder.) YouTube and others have taken to deplatforming and demonetizing certain users for purported community guidelines violations. Being privately-held enables YouTube to “erase” individuals from it site, many of who receive substantial compensation via ads and links to merchandise and personal web sites. Further, eliminating or refusing to recommend videos because its content is “harmful” or “offensive” or “inaccurate” discredits the source and/or compromises their reputations.

Some argue that Google’s status as a privately held company is purely theoretical. A growing body of court decisions have found that ISPs such as Twitter should be regarded as virtual public forums. If this is the case, then the First Amendment would arguably recognize that all forms of speech should be protected equally, hence precluding Google’s search results from favoring sites based on ideology.

On its face, using potentially anticompetitive means to engage in content and viewpoint discrimination appears illegal. Although Google’s status as a private corporation exonerates it from First Amendment restrictions, opponents argue that selective suppression of conservatives constitutes unfair business practices. However, if we subscribe to Volokh’s contentions, then not only is Google immune from legal culpability for any perceived censorship on its part under First Amendment laws— regardless of whether it is regarded as a private entity or public forum — but the threat of antitrust enforcement is also jeopardized insofar as such action arises out of or proximately relates to Google’s protected speech.

President Trump is perhaps one of media’s and Big Tech’s biggest adversaries. During his tenure as Commander-in-Chief, he has rhetorically lambasted companies such as Google, The New York Times, and Twitter for promoting “fake news” as well as censoring conservatives and far-right users, which in turn is creating an echo chamber of Leftist and Liberal-leaning dogma.

In August 2018, Trump tweeted that “Google search results for ‘Trump News’…[are] ‘rigged.’” He further added, “Google & others are suppressing voices of Conservatives and hiding information and news that is good.” While there is compelling evidence to support Trump’s claims, again, if we give deference to Volokh’s theories, then any bias on Google’s behalf, real or perceived, is permitted by law.

Superseding the President’s tweet, in December 2018, Google was called to testify before the House Judiciary Committee over concerns of political bias, data collection policies, and its growing presence in China. In response to questions pertaining to Google’s purported bias against conservatives in its search results and across its YouTube service, Google CEO Sundar Pichai testified, “I lead this company without political bias and work to ensure that our products continue to operate that way.”

Critics have since questioned the validity of Pichai’s testimony. The Daily Caller has recently published stories disclosing information contained in internal documents evidencing Google’s creation of internal blacklists to alter search results. In April 2019, the Caller exposed “The deceptive_news domain blacklist,” which is designed to ban certain sites — including several conservative-based sites — from appearing in any search feature or news product that violates Google’s Trust & Safety Team policies. Otherwise known as the “misrepresentation policy” and the “good neighbor policy,” the aforementioned blacklist, according to the article, may not impact the “ten blue links” but does affect all other forms of searches.

Additionally, internal documentation indicates that the blacklist is maintained using a “manual review tool,” which, if valid, infers that humans are responsible for controlling this feature and appears to contradict Mr. Pichai’s proclamation before the House: “Our [Google’s] algorithms have no notion of political sentiment.”

The second article circulated on June 11, 2019, elucidates two additional blacklists, one that prohibits “featured snippets” (also known as answer blocks) that violate Google policies from appearing in search results. The second, entitled “all_fringe_domain,” appears to be an algorithm that weeds out what the article calls “inappropriate web answers.” Breitbart and the American Spectator are both cited on this list.

Notwithstanding any perceived First Amendment rights, both articles illustrate how Google is developing algorithms designed to suppress certain types of information and opinions based on lists compiled by its (human) employees. Clearly these computerized programs are specifically designed to bias search results and thus discriminate against certain political content and viewpoints.

Yet assessing the above-mentioned protocols in the context of Volokh’s findings presents a conundrum: How does Google’s reliance on human judgment to manage its search engine results coincide with the provisions of Section 230 which is supposed to prohibit ISPs from altering third-party content?

Section 230(c)(1) of CDA reads:

“No provider or user of an interactive computer service shall be treated as a publisher or speaker of any information provided by another information content provider.”

Based on the foregoing, the rudimentary distinction between an ISP and a publisher is the former only provides access to information, it does not alter or edit content as in the case of a publisher. But an innate aspect of Google’s business model is having to select and rank information provided in search results based on the company’s editorial discretion. Human judgment is therefore required to both develop search-based algorithms and to facilitate their use.

Primarily, the First Amendment fully protects search engine results. In Search King, Inc. v. Google Technology, Inc., the federal court held that Google’s page rankings were “subjective result[s]” and therefore considered to be “constitutionally protected opinions.” Moreover, citing several precedent cases, in Langdon v. Google, Inc., the court also refused to order Google and Microsoft to give deferential treatment to Plaintiff’s site in their search results. Plaintiff’s request for injunctive relief was denied because it sought to “contravene[s] Defendants’ First Amendment rights.” The holding was premised on the previous rulings wherein newspapers could not be forced to print content or advertising; thus, search engines could not be forced to include links.

Google’s judgment as to what constitutes “hate” speech and corresponding use of community guidelines to compromise conservatives and dissident media is particularly egregious considering that the Supreme Court has made clear that there is no hate speech exception to the First Amendment. We can derive from this behavior that Google’s impaneled “judges” (i.e. executives) have created a de facto government intent to create its own set of laws. However, such governance on Google’s part, albeit alarming, is founded on the same principles as those that prevent the Courts from establishing a legally distinct category for “hateful content.”

Just as the High Court held that “[s]peech may not be banned on the ground that it expresses ideas that offend” (Justice Samuel Alito, Matal v. Tam (2017)), it also recognizes that the Constitution is intended to protect editorial decisions made by the press, even when it shields “…choices of content that in someone’s eyes are misguided, or even hurtful.” (Hurley, 515 U.S. at 574.)

Even if its decision-making process demonstrates a pattern of bias and is consequently harmful to certain sites, Google’s manipulation of search engine results is legally indistinct from The New York Times deciding what news to feature and how it should be arranged. Lovell v. City of Griffin, 303 U.S. 444, 452 (1938) found that freedom of the press is inclusive of “…every sort of publication which affords a vehicle of information and opinion,” which applied in today’s society extends to Google and other search engines.

Perhaps most importantly, the Courts recognize that human judgment is quintessential to processes used by Google and the Times. Collating search results and news articles, respectively, is inherent to the services these entities provide. So even in instances when a corporation asserts that its methodologies are based on objectivity, it would be impossible to extract the human assessment aspect of discerning what content should or should not be included.

Furthermore, Section 230 of the CDA additionally bars ISPs from civil liability for voluntarily restricting access to material the provider or user considers to be “obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable, regardless of whether the material is constitutionally protected. (See CDA Section 230(c)(2)).

Overall, the CDA affords ISPs great discretionary liberties when it comes to disseminating information. Under the statute, these platforms presumably cannot modify user content, only offer access to the same. This restriction distinguishes an ISP from a publisher. Yet it also permits ISPs to screen third-party content and remove material it deems to be objectionable from its platform. Such discernment requires editorial judgment and consequently allows ISPs to edit information. This capability seemingly contravenes the purpose of Section 230.

Section 230 additionally does not state that search engine ISPs such as Google are required to include all sources of information. Such a prerequisite requirement would be an impossible task. Nor does it dictate how Google should compile the third-party information it selects to feature in its search results. Similar to traditional publishers, Google is free to select and how to compile the material it references.

Moreover, the conflicting terms of Section 230 further supports Google’s legal right to adjust search engine results based on its subjective criteria.

Based on the foregoing, Eugene Volokh makes a hell of a convincing case in support of Google’s constitutional right to prioritize certain search results over others, even if the mechanisms it utilizes to do so are premised on its political biases. Furthermore, the language of Section 230 of the CDA additionally shields Google from civil liability for promoting third-party information which supports such bias.

Considering First Amendment case precedent and statutory tech laws both appear to enable Google’s arguably dubious business practices, does the DoJ’s threat of regulatory intervention have merit? Or is antitrust regulation useless in this scenario and merely a veiled threat? Can our current laws support government intervention and pressure Google into altering its business practices to protect against censorship? The answers to these questions are about as complicated as finding a way to reign in and oversee the activities of those who manage Big Tech.

In light of Volokh’s White Paper, it seems that our Justice Department will need to continue its search for laws that can defend Americans from being governed by the growing dictatorship of Big Tech. If this fails to happen, then Google’s right to free speech might just silence us all.