close
close

Eleventh Circuit Clarifies Scope of Abstraction-Filtration-Comparison Test for Copyright Claim

“Although the district court considered the selection and layout of the Compulife code to some extent, the district court never identified the entire layout of these variables in the code as a component of the code.” – Court of Appeals for the Eleventh Circuit

Eleventh CircuitOn Thursday, the United States Court of Appeals for the Eleventh Circuit released its opinion in the case Compulife Software, Inc. v. Newman, a case of copyright infringement and misappropriation of trade secrets that has been pending in the courts for some time.

The original complaint alleged that the defendants, consisting of multiple individuals, obtained access through improper means to data used by Compulife in its software. The software at issue allowed users to calculate personalized life insurance policy quotes using publicly and privately obtained data from various providers. Through various means, including a supervised scraping attack on Compulife’s website, the complaint alleged that the defendants obtained access to millions of quotes generated by its software and implemented the quotes on their own websites, thereby reducing Compulife’s sales.

Both sides waived their right to a jury trial, and on the original appeal, the Eleventh Circuit sent the case back for a new trial, “clarifying” some of the legal issues at issue. That decision, called Compulife I by the Eleventh Circuit, was discussed in a two-part article published on IPWatchdog upon its release in 2020.

At the second trial, a new judge found that Compulife’s copyright claim was invalid because most of its code was unprotectable. Applying the original appellate decision that scraping Compulife’s website constituted “proper means” for asserting a trade secret misappropriation claim, the district court granted an injunction while awarding damages and punitive damages to Compulife against all defendants, jointly and severally.

Both sides appealed, with Compulife arguing that the court erred in finding there was no copyright infringement, while the defendants argued that there was no trade secret misappropriation and, even if there had been, they should not be held jointly and severally liable.

Copyright infringement

The Court of Appeal first considered Compulife’s appeal of the copyright infringement claim. If it could be established that actual copying as well as legal office, Compulife could then have a valid infringement claim that would justify referral. Compulife I The court had already found actual copying, so it used the abstraction-filtration-comparison test to determine whether legal copying had occurred.

First, the opinion broke down the allegedly infringed program into its component structural parts. It agreed with Compulife that the arrangement of its source code, although a literal element of the program and therefore arguably distinguishable from other cases in which only nonliteral elements were considered, was a potentially protectable component part of the program. The district court looked at the individual parts but did not look at the arrangement as all. The court of appeal explained:

“Although the district court gave some consideration to the selection and layout of Compulife’s code, it never identified the entire arrangement of these variables in the code as a constituent element of the code.”

The case was therefore referred back for retrial.

The opinion then addressed whether the district court properly sifted out all of the ineligible material. While Compulife argued that the district court found no protection in several elements, the appellate court disagreed, affirming instead that the use of certain variables was “too obvious” to be an original work, among other individual findings.

Finally, the opinion compared the protected material with the alleged imitator to determine whether both the quantitative and qualitative significance of the copied quantity justified a finding in Compulife’s favor. The court of appeals, explaining that the arguments concerning system code to be established in the course of the proceedings, disagreed with Compulife’s argument regarding the remaining elements of the code – while 84% of Compulife’s code was copied, only about 8% of it under protection code was copied. Furthermore, the appellate court could find no significance in the remaining elements, meaning that Compulife’s argument failed to establish both the quantitative and qualitative significance of copying.

Misappropriation of Trade Secrets and Liability

The appellate court then considered whether the trial court had erred in its ruling on the misappropriation of trade secrets. The court in Compulife I the existence of a trade secret has already been established, and the appellate court once again had no problem ruling that the defendants used improper means to obtain the trade secret, even if those means were not “independently unlawful”:

“…scraping and related technologies… can be completely justified. Much of the modern Internet relies on these technologies. But the defendants in this case did not take innocent screenshots of a publicly accessible website; instead, they copied the sequence of Compulife’s copyrighted code and I used this code “to conduct a scraping attack that resulted in access to millions of variable-dependent insurance quotes.”

As a result, the appellate court found no error in the district court’s ruling on the trade secret misappropriation. Furthermore, the appellate court found no error in finding that the defendants were jointly and severally liable because they acted “in concert” and that trade secret misappropriation is an intentional tort that falls within the exception to Florida’s comparative fault regime.

Image Source: Deposit Photos
Author: AndreyPopov
Image ID: 211210396

Picture by Lucas Thrun