Draft website · details subject to confirmation

1st Workshop on Privacy-Preserving Artificial Intelligence in Regulated and Cyber-Physical Systems
(PrivAI-RCPS) 2027

28 June–1 July 2027 · Egham, United Kingdom

Co-located with the 25th International Conference on Applied Cryptography and Network Security (ACNS 2027)

Full-day, in-person workshop · exact workshop date to be announced

Campus photograph: ACNS 2027 website

Call for Papers

Important Dates

  • Paper submission deadline: 19 February 2027 (suggested; provisional)
  • Notification to authors: 21 March 2027 (suggested; provisional)
  • Camera-ready papers: 14 June 2027 (ACNS proceedings deadline)

Submission and notification dates are provisional until confirmed by the workshop chairs. The workshop’s exact day within 28 June–1 July 2027 and deadline time zones will be announced. ACNS Conference Page↗

Overview

Artificial intelligence is increasingly embedded in cyber-physical systems operating in healthcare, manufacturing, transportation, energy and critical infrastructure. These systems process sensitive personal, operational and sensor data while interacting with physical environments and making decisions with potential safety and societal consequences.

Federated learning, differential privacy, secure computation, confidential computing and synthetic data offer promising ways to use sensitive data without centralising or directly disclosing it. Their deployment must also account for safety, latency, communication, resource, auditability and reliability constraints.

PrivAI-RCPS brings together researchers and practitioners to advance the design, evaluation, attack analysis, assurance and practical deployment of private and secure AI in sensitive and safety-critical settings.

Topics of interest

Topics include, but are not limited to, the following:

Privacy-preserving distributed AI

  • Federated, split, decentralised, and collaborative learning
  • Differential privacy and privacy accounting
  • Secure aggregation, secure computation, and confidential computing
  • Privacy-preserving edge intelligence
  • Communication- and resource-efficient private learning

Privacy attacks and adversarial security

  • Membership, attribute, reconstruction, and model-inversion attacks
  • Privacy leakage through gradients, model updates, and outputs
  • Poisoning, backdoor, and malicious-client attacks
  • Defences for adversarial or compromised distributed environments
  • Threat modelling for AI-enabled cyber-physical systems

Generative AI and synthetic data

  • Privacy-preserving generative models
  • Federated training or adaptation of foundation models
  • Synthetic healthcare, industrial, and sensor data
  • Disclosure and re-identification risks in synthetic data
  • Private retrieval-augmented generation and machine unlearning

Regulated and cyber-physical applications

  • Healthcare, medical imaging, and the Internet of Medical Things
  • Industrial IoT, operational technology, and smart manufacturing
  • Intelligent transportation and connected vehicles
  • Energy systems, smart grids, and critical infrastructure
  • Robotics, wearables, autonomous systems, and privacy-aware sensing

Assurance and evaluation

  • Formal and empirical evaluation of privacy guarantees
  • Privacy–utility–fairness and privacy–safety trade-offs
  • Auditability, traceability, and continuous monitoring
  • Evaluation under non-IID data and distribution shifts
  • Reproducible benchmarks, testbeds, and real-world deployment studies

Submission Guidelines

Submission categories

  • Regular research papers presenting mature technical contributions.
  • Short papers presenting emerging ideas or preliminary results.
  • System and experimental survey papers describing implementations or deployments.
  • Dataset, benchmark and artifact papers supporting reproducible evaluation protocols.

Submissions will undergo rigorous peer review, with a minimum of three independent reviews as proposed. Evaluation will consider technical quality, originality, relevance, clarity and potential to advance the workshop community.

Proceedings and attendance

Accepted papers will appear in the joint ACNS workshop post-proceedings published in Springer’s LNCS series, subject to the conference’s publication requirements. Presentations are expected to be delivered in person. At least one author of each accepted paper must register as required by ACNS.

The Microsoft CMT service was used for managing the peer-reviewing process for this conference. This service was provided for free by Microsoft and they bore all expenses, including costs for Azure cloud services as well as for software development and support.

Programme

The workshop programme will be announced here once confirmed.

Committee

Workshop Chairs

Dr Peng Liu

Department of Mathematical Sciences
Loughborough University, United Kingdom

Dr Akarsh K. Nair

Department of Computer Science
Nottingham Trent University, United Kingdom

Programme Committee

  • Prof Huiyu Zhou — University of Leicester, UK
  • Prof Victor Chang — Aston University, UK
  • Dr Lu Zhang — Loughborough University London, UK
  • Dr Weiwei Jiang — Beijing University of Posts and Telecommunication
  • Dr Miaomiao Yu — East China Normal University
  • Dr Deepti Gupta — Texas A&M University, USA
  • Dr Sachin DN — IIIT Bhopal, India