The rapid growth of machine learning as a service has revolutionized industries such as healthcare, finance, transportation, and e-commerce. While ML significantly enhances service efficiency and personalization, it raises considerable concerns regarding privacy breach and unauthorized data exploitation. Thus, ensuring privacy in ML-driven services has become critical for gaining public trust and the compliance with data regulations.
Privacy-preserving machine learning (PPML) is attracting increasing attention. This paradigm resorts to advanced privacy-enhancing techniques (PET) such as secure multi-party computation (SMPC), homomorphic encryption (HE), differential privacy (DP), and trusted execution environments (TEE), to protect both ML models and sensitive data throughout the ML lifecycle. However, delivering scalable, efficient, and secure ML services while preserving privacy remains challenging due to issues such as computational complexity, data heterogeneity, real-time constraints, and model accuracy trade-offs.
This special issue aims to bridge research gaps at the intersection of machine learning, service computing, and privacy-enhancing techniques. We invite high-quality original submissions addressing theoretical advances, practical algorithms, frameworks, and methodologies that leverage PETs to secure ML service in a privacy-preserving, efficient, and robust manner.
Topics of Interest
Topics of interest include, but are not limited to:
- Federated learning for privacy-preserving ML services
- Secure multi-party computation and homomorphic encryption for ML-as-a-service
- Differential privacy methodologies and their integration into ML services
- Trusted execution environments for secure and private ML computations
- Privacy-preserving techniques in distributed ML service architectures (e.g., edge computing)
- Scalable privacy-preserving ML protocols for real-time analytics and streaming data services
- Privacy-preserving deep learning for computer vision, natural language processing, and multimodal service applications
- Privacy threat analysis in AI agent services
- Privacy-assured solutions for securing AI agent service and service protocols
- Privacy-preserving recommendation services and personalized service delivery
- Privacy-preserving ML-based analytics in real-world sectors such as healthcare, financial services, transportation, and smart city infrastructures
- Fairness, transparency, interpretability, and trust in privacy-preserving ML services
- Benchmarking, metrics, and evaluation methodologies for privacy-preserving ML service systems
- Privacy risk mitigation in PPML service deployment
- Privacy risk assessment, auditing, and compliance in ML services
Important Dates
- Manuscript Submission Deadline: October 1, 2025
- First Round Notification: February 1, 2026
- Revised Manuscript Due: March 10, 2026
- Final Decision Notification: April 20, 2026
- Final Manuscript Submission Due: April 30, 2026
- Expected Publication: Mid 2026
Guest Editors
- A/Prof. Xingliang Yuan, The University of Melbourne, Australia
- Prof Ronald Cramer, CWI and Leiden University, Netherlands
- Prof Kwok Yan Lam, NTU, Singapore
- Dr Maggie Liu, RMIT University, Australia
Submission Guidelines
For author information and guidelines on submission criteria, please visit Author Resources . Authors should submit original manuscripts not exceeding 14 pages following IEEE Transactions on Services Computing guidelines. All submissions must be made through the IEEE Author Portal. Please select “Special Issue on Privacy-Preserving Machine Learning Services” during submission. Manuscripts must not be published or under review elsewhere, and should provide at least 30% original technical contributions compared to related publications.
In addition to submitting your paper to IEEE Transactions on Services Computing, you are also encouraged to upload the data related to your paper to IEEE DataPort. IEEE DataPort is IEEE’s data platform that supports the storage and publishing of datasets while also providing access to thousands of research datasets. Uploading your dataset to IEEE DataPort will strengthen your paper and will support research reproducibility. Your paper and the dataset can be linked, providing a good opportunity for you to increase the number of citations you receive. Data can be uploaded to IEEE DataPort prior to submitting your paper or concurrent with the paper submission.
