Cognizant AI Lab Achieves 59th U.S. Patent with New Innovations

Cognizant has announced that its AI Lab has been granted two new U.S. patents, bringing the total to 59, with 23 more patents pending. In a press release, Cognizant highlighted the significance of these patents, which include innovations in deep neural network architectures and optimizing loss functions for AI models.

The first patent, issued on April 22, 2025, covers a method for the multi-objective coevolution of deep neural network architectures, aimed at enhancing model performance and resource efficiency. The second patent, issued on May 6, 2025, focuses on optimizing loss functions through Taylor Series Expansion to improve training efficiency and model robustness, particularly in data-limited scenarios.

In addition to these patents, Cognizant's AI Lab received a Gold Award at the Genetic and Evolutionary Computation Conference (GECCO) 2025 for its research on "Realizing Human Expertise through AI." This research, developed by Dr. Elliott Meyerson and Professor Risto Miikkulainen, employs evolutionary AI to refine decision strategies, surpassing the performance of individual human submissions in pandemic policy recommendations.