Deepcell and Gilead Sciences launched a co-development effort to build an AI foundation model for Chinese hamster ovary (CHO) cell line development using cell morphology. The partnership will train a custom model using Deepcell’s REM-I platform data and self-supervised learning to characterize morphological structure without manual labeling. The model is intended to provide earlier readouts of cell line quality—such as titer, stability, aggregation risk, and metabolic profile—during cell line development, potentially improving the speed of candidate selection and reducing time spent on manual interpretation. Deepcell said it expects to publish an open-weight version of the initial model for broader research and bioprocessing use. For biologics manufacturers, the deal highlights where AI investment is moving next: not only discovery and assay analysis, but also upstream bioprocess decision-making tied to CHO performance and production robustness.