Publicly Available Data and the Misappropriation Doctrine
This blog post examines an ongoing legal dispute over access to publicly available internet data between HiQ Labs and LinkedIn Corp. First, it recounts the facts and procedural posture of the case. Second, it describes the case law and history of the misappropriation doctrine, a legal doctrine that could prove decisive in the dispute. Third, taking the law and policy goals of the misappropriation doctrine into account, this blog post analyzes the dispute, and determines how a court should ultimately rule.
HiQ Labs v. LinkedIn
Since 2017, LinkedIn and hiQ have been involved in a legal dispute regarding hiQ’s access to LinkedIn’s publicly available data. See HiQ Labs, Inc. v. LinkedIn Corp., 273 F.Supp.3d 1099, 1104 (N.D. Cal. 2017) (Chen). LinkedIn is a social network that allows its users to post their professional work information — current occupation, education history, and previous work experience, etc. — on a public platform. HiQ Labs, Inc. v. LinkedIn Corp., 938 F.3d 985, 990 (9th Cir. 2019) (Wallace, Berzon, Berg) aff’g and remanding 273 F.Supp.3d 1099 (N.D. Cal. 2017) (Chen). HiQ is a data analytics company that uses software to acquire massive amounts of publicly available data from websites like LinkedIn.com. Id.
HiQ uses this data for two purposes. Id. First, it can indicate to employers which employees are most at-risk for leaving, and this allows employers to efficiently allocate salaries and benefits to retain them. Id. Second, hiQ identifies skills gaps within companies. Id. This allows businesses to invest in training, which is apparently less expensive than recruiting external hires. Id. However, LinkedIn wants to monetize the vast amount of data it has accrued, and create products that are similar to those that hiQ produces. Id. at 991. Because of this, LinkedIn denied hiQ access to its publicly available data. Id.
HiQ sued LinkedIn, and was granted a preliminary injunction by the district court. The case was appealed to the Ninth Circuit, and LinkedIn based its argument primarily on the Computer Fraud and Abuse Act (CFAA), which prohibits access to computers without “authorization.” Id. at 996. In legal battles over publicly available data, there is a question as to whether the CFAA is an “anti-intrusion” statute or an “anti-misappropriation” statute. Id. Under the former, it would not apply to publicly available data, but under the latter, it would. Id. at 997. In this case, the Ninth Circuit held that the CFAA is an anti-intrusion statute, and rejected LinkedIn’s argument. Id. Upon remand, LinkedIn filed a counterclaim invoking the misappropriation doctrine. HiQ, slip-op. at 7.
The Misappropriation Doctrine
In the context of intellectual property law, one may own a copyright, trademark or patent, but one cannot acquire a right to facts. However, there are some cases where this latter point is not entirely true. Sometimes, one party has invested time and effort into acquiring information, and another party attempts to take advantage of the first party’s efforts. This is referred to as “misappropriation.”
The most seminal case in this field is International News Service v. Associated Press, 248 U.S. 215 (1918) aff’g245 F. 244 (2d Cir. 1917) (Ward, Rogers, Hough) modifying 240 F. 983 (S.D.N.Y. 1917) (Hand). During the First World War, AP would report on the conflict by publishing dispatches using information that it had independently acquired. William M. Landes & Richard A. Posner, The Economic Structure of Intellectual Property Law 105 (2003). INS, which had not independently acquired this information, would rewrite the dispatches that AP published on the East Coast of the United States, and publish them in newspapers on the West Coast. Id. The Supreme Court ruled that AP did have a “quasi-property” right to this information vis-a-vis INS. Id. at 222. In creating this quasi-property right, the Supreme Court emphasized the “expenditure of labor, skill, and money” that the AP had invested in acquiring the information, and found INS’s actions problematic for attempting “to reap where it has not sown.” Id. In other words, INS was attempting to free ride on AP’s work, and if it had been allowed to do this, AP would not have gained the rewards of its activity, and would be deterred from continuing this practice in the future. See Landes & Posner at 105.
INS was decided over one-hundred years ago, and because of Erie Railroad Co. v. Tompkins, 304 U.S. 64 (1938), it is no longer binding precedent, “having been based on the federal courts’ subsequently renounced authority to formulate common law principles in suits arising under state law though brought in federal court, usually under the diversity jurisdiction.” Id. Therefore, the misappropriation doctrine is no longer federal law, but it is valid under the common law of some states. In order to understand the misappropriation doctrine, there is still value in learning about INS because it created the doctrine, has a straightforward fact pattern, explains the logic behind it in a clear way, and some states even cite INS verbatim when adjudicating misappropriation claims. HiQ, slip-op. at 8. Misappropriation suits do not have to deal in “news” as such. See National Basketball Association v. Motorola, Inc., 105 F.3d 851 (2d Cir. 1997) (Van Graafeildand, Winter, Altimari) aff’g in part, vacating in part 939 F.Supp. 1071 (S.D.N.Y. 1996) (Preska). Take, for example, National Basketball Association v. Motorola, 105 F.3d 851 (2d Cir. 1997) (Van Graafeildand, Winter, Altimari) aff’g in part, vacating in part 939 F.Supp. 1071 (S.D.N.Y. 1996) (Preska), which was decided by the Second Circuit in 1997. In NBA, Motorola produced a pager product that would allow fans to get real-time information relating to NBA games, such as the score, time left, and individual player statistics. Id. The product was called “SportsTrax.” Id. The NBA had recently introduced its own similar product known as “Gamestats.” Id. The NBA sued under the misappropriation doctrine, and argued that Motorola was free riding by monetizing a product that used information produced by the NBA. Id. at 846. The Second Circuit considers multiple elements in a misappropriation claim, one of which is free riding. See Landes & Posner at 109. In particular, the Second Circuit wrote that, “[a]n indispensable element of an INS … claim is free riding by a defendant on a plaintiff’s product, enabling the defendant to produce a directly competitive product for less money because it has lower costs.” NBA, 105 F.3d at 853.
In the end, the Second Circuit held that Motorola was not free riding by producing its pager. Id. at 853. First, the NBA’s primary product is the broadcast of basketball games, and the pager in no way competed with this product. Id. Second, insofar as Motorola’s SportsTax competed with the NBA’s Gamestats, this did not occur on the basis of free riding. Id. Motorola spent its own time and effort collecting information from NBA games, and did not take advantage of the NBA’s time and effort in doing the same. Id. Ultimately, the court held, if the NBA was going to create a product to compete with Motorola’s, this competition should occur on the basis of the technical quality and cost to consumers of the two products. Id.
Analysis of HiQ Labs v. LinkedIn
The Northern District of California denied hiQ’s motion to dismiss LinkedIn’s misappropriation claim. Id. at 8. The elements that California considers in a misappropriation analysis are as follows: “(1) the plaintiff has invested substantial time and money in development of its … ‘property’; (2) the defendant has appropriated the [property] at little or no cost; and (3) the plaintiff has been injured by the defendant’s conduct.” Balboa Ins. Co. v. Trans Global Equities, 218 Cal.App.3d 1327, 1334 (1990). These elements ask a court to determine whether or not the defendant is free riding vis-a-vis the plaintiff. Ultimately, the court wrote, “there are questions of fact as to whether hiQ is a free rider with respect to LinkedIn and whether hiQ’s use of LinkedIn’s data reduces LinkedIn’s incentive to invest in its infrastructure.” HiQ, slip-op. at 8.
Although previous cases have addressed other contexts in which one party took advantage of the information acquired by another, this case would be among the first to determine the kind of protection that publicly available internet data is afforded. The scale of the information in this case dwarfs that of previous cases. Whereas NBA involved the information produced by a few basketball games, LinkedIn has 500 million users, and hiQ wants their data. Id.Because of this, the economic implications of a court ruling one way or another may be much greater than in previous misappropriation cases. If a court hears this case, how should it decide?
Of course, the Second Circuit does not have binding precedent over the Ninth Circuit. And it should be noted that the Second Circuit considers additional elements in misappropriation claims that California does not explicitly consider, such as whether the information in question is time sensitive, and whether “the defendant is in direct competition with a product or service offered by the plaintiffs.” NBA, 105 F.3d at 846. However, for the purposes of analyzing the free riding element — which is all that California considers — NBA could hardly be more similar to the suit involving hiQ and LinkedIn.
Like Motorola, hiQ exerted its own time and effort gaining data from LinkedIn. Like the NBA, LinkedIn’s primary product exists separately from its secondary product. That is to say, LinkedIn’s social network does not depend on the existence of products that monetize the data accumulated by the social network. If a company like hiQ comes along and produces data-using products that outcompete LinkedIn’s data-using products, this would not diminish LinkedIn’s value as a social network. Contrast this with INS. If INS had been allowed to copy AP’s dispatches, then AP could not have continued its service — INS was producing a highly substitutable product, but AP was the only company of the two that incurred serious costs in doing so. If AP had ceased this particular service because of free riding, the consumers that wanted it would have wound up with nothing. On the other hand, LinkedIn exists primarily as a social network. HiQ may obtain data from LinkedIn, but it is not producing a competing or substitutable social network.
Therefore, regardless of hiQ’s actions, consumers will retain LinkedIn’s value as a social network, but as far as data-related products go, consumers would be best served if LinkedIn’s data-using products compete with hiQ’s data-using products on the basis of quality and price. For this reason, a court should not find hiQ civilly liable for misappropriation.