Recently, Ant Group Digital Technology has achieved a notable milestone, with two of its cutting-edge privacy computing technology innovations being selected for presentation at the ACM CCS—a premier global security conference—and for publication in IEEE TDSC, a leading journal in the field. This accomplishment reaffirms Ant Group Digital Technology's sustained leadership in privacy-preserving artificial intelligence technology. The two studies center on the GBDT model, a widely utilized tool in cross-institutional collaborative modeling. They specifically tackle the dual challenges of safeguarding data privacy and achieving high-performance computing in joint modeling and inference scenarios. At present, the industry predominantly relies on federated learning, a method that inherently carries risks of information leakage. In contrast, Ant Group Digital Technology has opted for the multi-party secure computation technology pathway, attaining dual breakthroughs in both security and efficiency during the training and inference phases. Currently, the fruits of this research have been integrated into Ant Group Digital Technology's privacy computing products, facilitating secure cross-institutional data collaboration. Both ACM CCS and IEEE TDSC are renowned for upholding the highest academic standards in their respective domains.