Abstract
Like a hamster wheel, the “virtuous cycle” of data collection, collation, and computation spins to create algorithms to assist modern business operations. Companies accumulate data by the terabyte with hopes that their army of data scientists can make sense of all the noise using machine learning. Potential profits encourage companies to seek out data—sometimes unlawfully through collusion. The use of unlawfully collected data in algorithms harms competition. Such data provides a competitive edge that is not accessible to all marketplace participants. Antitrust laws can police unlawful data sharing, but existing remedies are not effective for restoring competition.
Merely deleting ill-gotten data does not eliminate competitive harm. Generated algorithmic outputs are derived from the training data used to create the model. Algorithmic deletion, or compelled deletion of an algorithm, is a necessary remedy. By being forced to delete an algorithm, companies cannot benefit from unlawful conduct. Algorithmic deletion also provides a simple remedy to a complicated technological problem. The FTC’s recent consumer protection enforcement has successfully used algorithmic deletion in recent years and should now be a consideration when dealing with antitrust violations. This Article will argue for why algorithmic deletion is an appropriate remedy as artificial intelligence becomes embedded in the economy.
Recommended Citation
Vaibhav Srikaran,
BEYOND DATA DELETION: ADDRESSING ANTICOMPETITIVE CONDUCT IN THE ERA OF MACHINE LEARNING,
20 Wash. J. L. Tech. & Arts
(2025).
Available at:
https://digitalcommons.law.uw.edu/wjlta/vol20/iss2/4