terms of use

Subscribe to terms of use
  • Flight and travel data has always been valuable for data aggregators and online travel services and has prompted litigation over the years.
  • Latest suit from Air Canada against a rewards travel search site raises some interesting liability issues under the CFAA.
  • The implications of this case, if the plaintiffs are successful, could impact the legal analysis of web scraping in a variety of circumstances, including for the training of generative AI models.

In a recent post, we recounted the myriad of issues raised by recently-filed data scraping suits involving job listings, company reviews and employment data.  Soon after, another interesting scraping suit was filed, this time by a major airline against an award travel search site that aggregates fare and award travel data.  Air Canada alleges that Defendant Localhost LLC (“Localhost” or “Defendant”), operator of the Seats.aero website, unlawfully bypassed technical measures and violated Air Canada’s website terms when it scraped “vast amounts” of flight data without permission and purportedly caused slowdowns to Air Canada’s site and other problems. (Air Canada v. Localhost LLC, No. 23-01177 (D. Del. Filed Oct. 19, 2023)).[1]   

The complaint alleges that Localhost harvested data from Air Canada’s site and systems to populate the seats.aero site, which claims to be “the fastest search engine for award travel.” 

It also alleged that in addition to scraping the Air Canada website, Localhost engaged in “API scraping” by impersonating authorized requests to Air Canada’s application programming interface.  

Many online services feature comprehensive terms of use intended to protect their business from various types of risks.  While it is often the case that a great deal of thought goes into the creation of those terms, frequently less attention is paid to how those terms are actually presented to users of the service. As case law continues to demonstrate, certain mobile and website presentations will be held to be enforceable, others will not.  Courts continue to indicate that enforceability of terms accessible by hyperlink depends on the totality of the circumstances, namely the clarity and conspicuousness of the relevant interface (both web and mobile) presenting the terms to the user. In a prior post about electronic contracting this year, we outlined, among other things, the danger of having a cluttered registration screen.  In this post, we will spotlight five recent rulings from the past few months where courts blessed the mobile contracting processes of e-commerce companies, as well as one case which demonstrates the danger of using a pre-checked box to indicate assent to online terms.

The moral of these stories is clear – the presentation of online terms is essential to enhancing the likelihood that they will be enforced, if need be. Thus, the design of the registration or sign-up page is not just an issue for the marketing, design and technical teams – the legal team must focus on how a court would likely view a registration interface, including pointing out the little things that can make a big difference in enforceability. A failure to present the terms properly could result in the most carefully drafted terms of service ultimately having no impact on the business at all.

In recent years, courts have issued a host of rulings as to whether online or mobile users received adequate notice of and consented to user agreements or website terms when completing an online purchase or registering for a service. Some online agreements have been enforced, while others have not. In each case, judges have examined the circumstances behind the transaction closely, scrutinizing the user interface and how the terms are presented before a user completes a transaction. In general, most courts seek to determine whether the terms are reasonably conspicuous to the prudent internet user and whether the user manifested sufficient assent by signing up for a service or completing a transaction.

From the perspective of making a sign-up process as smooth as possible, there is often an interest in moving the reference to terms and conditions out of the main flow of user sign-ups.  However, as we were reminded recently by an Illinois court examining the interfaces of DVD rental company Redbox, one does so at risk of finding those terms to be unenforceable.

The Illinois court noted numerous shortcomings with Redbox’s electronic contracting process.  It found that because links to the relevant terms were not clearly and conspicuously displayed, customers did not have constructive notice that they were assenting to those terms when hitting the “Pay Now” button to rent a DVD at a kiosk or by signing into a Redbox account online. (Wilson v. Redbox Automated Retail, LLC, No. 19-01993 (N.D. Ill. Mar. 25, 2020)). As such, the court denied Redbox’s motion to compel arbitration of plaintiff’s claims.

A recent dispute between an advertiser AXTS Inc. (“AXTS”) and a video production company GY6vids (“GY6”) produced an interesting issue involving the federal Computer Fraud and Abuse Act (CFAA) – that is, whether an entity that allegedly overloaded another company’s YouTube channel content with a flood of “dislikes” following a contractual dispute is liable under the CFAA for accessing a protected computer “without authorization.”  (AXTS Inc. v. GY6vids LLC, No. 18-00821 (D. Ore. Oct. 24, 2018)).

We have been closely monitoring the evolving state of the law regarding CFAA liability for certain commercial web scraping and related practices.  The instant case between AXTS and GY6 is a little different in that the claim did not arise from AXTS’s alleged access to video content stored on GY6’s network, but publically-accessible videos stored on a third-party’s (e.g., YouTube) servers.   

This past week, the Supreme Court denied the petitions for certiorari in two noteworthy Ninth Circuit decisions that had interpreted the scope of liability under the federal Computer Fraud and Abuse Act (CFAA) in the context of wrongful access of company networks by employees and in instances involving unwanted data…

UPDATE: On January 18, 2019, the Ninth Circuit affirmed the award of damages and injunctive relief in favor of Facebook. (Facebook, Inc. v. Power Ventures, Inc., No. 17-16161 (9th Cir. Jan. 18, 2019) (unpublished)). The California district court in 2017 had awarded Facebook almost $80,000 in CFAA damages, representing only the period after Facebook sent its cease and desist letter to the defendant and including expenses both for technical measures to block Power Ventures from accessing Facebook servers and expenses for negotiating with Power Ventures to voluntarily stop its activities and destroy the data.  The lower court also granted Facebook’s request for a permanent injunction barring defendant from, among other things, accessing Facebook for a commercial purpose without permission.

  • Unauthorized Access: A former employee, whose access has been revoked, and who uses a current employee’s login credentials to gain network access to his former company’s network, violates the CFAA. [U.S. v. Nosal, 2016 WL 3608752 (9th Cir. July 5, 2016)]
  • Data Scraping: A commercial entity that accesses a public website after permission has been explicitly revoked can be civilly liable under the CFAA. However, a violation of the terms of use of a website, without more, cannot be the basis for liability under the CFAA, a ruling that runs contrary to language from one circuit level decision regarding potential CFAA liability for screen scraping activities (See e.g., EF Cultural Travel BV v. Zefer Corp., 318 F.3d 58 (1st Cir. 2003)). [Facebook, Inc. v. Power Ventures, Inc., No. 13-17102 (9th July 12, 2016)]

This past week, the Ninth Circuit released two important decisions that clarify the scope of liability under the federal Computer Fraud and Abuse Act (CFAA), 18 U.S.C. § 1030.  The Act was originally designed to target hackers, but has lately been brought to bear in many contexts involving wrongful access of company networks by current and former employees and in cases involving the unauthorized scraping of data from publicly available websites.