By: Garrett Lipscomb
One of the most interesting aspects of Formula 1 (F1) is how every single detail can be interpreted as affecting a race outcome. After a race, people can ask endless questions about whether the driver’s skill, the team’s unique race strategy, or the car’s expertly engineered parts won the race. Imagine asking AI to compare years of statistics and results to help answer a question like this. For teams, this information could help them gain a competitive advantage over other teams who may not be able to afford the research. For fans, it could make the sport easier to understand and potentially grow the sport’s popularity. However, F1’s website restricts the use and commercial exploitation of its content.[1] Can Formula One retain control over this type of information as AI continues to grow?
The first problem for F1 is that a race result is a fact. In Feist Publications, Inc. v. Rural Telephone Service Co., the Supreme Court ruled that facts, alone or in a compilation, are not original and cannot be copyrighted.[2] However, a factual compilation featuring original selection or arrangement can be eligible for copyright.[3] For curious fans, or businesses looking to train their models to provide valuable information to F1 teams, that distinction matters. Copyright may protect an original arrangement of the race data while still allowing others to use the underlying results to develop their own analysis.
This type of conflict between a professional sports league’s commercial interest and the use of sports facts is not new. In National Basketball Association v. Motorola, Inc., the NBA challenged that Motorola and STATS could not send basketball scores and statistics to handheld pagers.[4] The Second Circuit held that neither company infringed NBA’s copyright because they reproduced only facts from the broadcast, not the game description that constitutes the broadcast itself.[5] This case offers a helpful comparison as an AI business using F1 results can likewise distinguish the facts it uses from F1’s protected coverage of the sport. This suggests that F1 does not have absolute copyright ownership of the race results.
Taken together, Feist and Motorola show that F1 cannot use copyright to lock up the underlying race results. It is important to emphasize the distinction between using race facts—which copyright leaves free—and accessing F1’s website and information. Companies training AI models should check website terms because those terms could restrict the collection of data. F1’s website terms restrict extracting and commercially exploiting their website content.[6] F1 owns the copyrighted broadcast, but that ownership does not extend to the underlying facts the broadcast conveys. This allows fans and businesses to utilize AI in creative ways to gather and interpret race data, which can enable a better and more in-depth understanding of the sport.
[1] Legal Notices, Formula 1, https://www.formula1.com/ [https://perma.cc/BWD8-GTXL].
[2] Feist Publ’ns, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 344–45 (1991).
[3] Id. at 348.
[4] Nat’l Basketball Ass’n v. Motorola, Inc., 105 F.3d 841, 843 (2d Cir. 1997).
[5] Id. at 846–47.
[6] Legal Notices, supra note 1.
