AI-assisted security evaluation
Physical leakage, fault analysis, vulnerability discovery, hardware design, and the evaluation and optimization of countermeasures.
8th International Workshop · ACNS 2027
Connecting AI-driven hardware-security evaluation with the security of AI hardware and implementations.
In conjunction with the 25th International Conference on Applied Cryptography and Network Security (ACNS 2027).
About the workshop
AIHWS brings together researchers and practitioners from hardware security, cryptography, machine learning, electronic design automation, and systems security.
Physical leakage, fault analysis, vulnerability discovery, hardware design, and the evaluation and optimization of countermeasures.
Physical and microarchitectural threats to embedded AI, processors, GPUs, accelerators, and the integrity and confidentiality of AI computations.
Contribute your research
Artificial intelligence (AI) and hardware security are becoming increasingly interconnected. Machine learning and automated reasoning offer new ways to analyze physical leakage, detect vulnerabilities, optimize countermeasures, assess implementation security, and support hardware design and validation. Conversely, the hardware platforms used to train and deploy AI—including embedded processors, RISC-V systems, GPUs, AI accelerators, and edge devices—create new opportunities for side-channel analysis, fault injection, model extraction, and attacks on the integrity and confidentiality of AI computations.
The 8th International Workshop on Artificial Intelligence in Hardware Security (AIHWS 2027), held in conjunction with ACNS 2027, aims to bring together researchers and practitioners from hardware security, cryptography, machine learning, electronic design automation, and systems security. The workshop seeks original contributions that advance either the use of AI to evaluate and improve hardware security or the understanding and protection of AI systems against hardware-level threats. We especially encourage work that connects new AI techniques to realistic security models, reproducible evaluation, and constraints of deployed systems.
Topics include, but are not limited to:
We welcome theoretical, methodological, and experimental contributions, including rigorously evaluated tools, benchmarks, case studies, and negative results that clarify the limits of existing approaches. Submissions will be peer reviewed through a competitive selection process. Accepted papers will be presented in person at the workshop and included in the ACNS workshop post-proceedings published by Springer in the LNCS series, subject to the conference's publication requirements. Detailed submission formatting, page limits, and dates will be announced on this website.
Contributions should have a clear hardware, implementation, or physical-security component.
Timeline
| Paper submission deadline | To be announced |
|---|---|
| Author notification | To be announced |
| Camera-ready deadline | To be announced |
| Workshop date | To be announced |
| ACNS 2027 conference | 28 June – 1 July 2027 |
The exact workshop date and paper deadlines will be posted when finalized.
For authors
We welcome original theoretical, methodological, and experimental contributions, including rigorously evaluated tools, benchmarks, case studies, and negative results. Submissions will be peer reviewed through a competitive selection process.
Formatting requirements, page limits, and the submission system will be announced here.
Accepted papers will be presented in person and included in the ACNS workshop post-proceedings published by Springer in the LNCS series, subject to conference publication requirements.
Workshop organization
Radboud University, the Netherlands.
Temasek Laboratories, Nanyang Technological University, Singapore.
University of Zagreb, Faculty of Electrical Engineering and Computing, Croatia; Radboud University, the Netherlands.
Membership is subject to individual confirmation.
Workshop-specific contact details will be announced.
In-person format
The full-day workshop will feature peer-reviewed research presentations, invited keynote talks, and a moderated discussion on open problems and reproducible evaluation.
Indicative program only. The schedule and invited speakers are not yet confirmed and are subject to ACNS scheduling and the number of accepted papers.
| 09:00–09:15 | Welcome, objectives, and introduction |
|---|---|
| 09:15–10:00 | Invited keynote: AI-driven physical-security evaluation |
| 10:00–10:30 | Coffee break |
| 10:30–12:00 | Technical session I: AI for side-channel and fault analysis |
| 12:00–13:00 | Lunch |
| 13:00–13:45 | Invited keynote: Security of AI hardware and accelerators |
| 13:45–15:15 | Technical session II: AI hardware threats and defenses |
| 15:15–15:45 | Coffee break |
| 15:45–16:15 | Technical session III: Emerging methods and benchmarks |
| 16:15–16:50 | Moderated discussion: trustworthy evaluation, reproducibility, and open challenges |
| 16:50–17:00 | Closing remarks |
Venue & registration
AIHWS takes place in conjunction with ACNS 2027 in Egham, United Kingdom, during the conference period of 28 June – 1 July 2027.
For venue, accommodation, travel and registration information, please consult the official ACNS 2027 website.
Since 2020
AIHWS has been part of ACNS since 2020, bringing together the AI and hardware-security research communities.