Call for Papers

The Safe AI workshop solicits contributions including but not limited to the following topics in all areas of AI and its application:

Foundations and different learning settings

• Uncertainty modelling and treatment for safe AI

• AI Robustness and Reliability

• LLMs and AI agents hijacking and defense mechanisms 

• Safe Reinforcement Learning

• Safe Federated Learning 

• Safe Frugal AI

• Safety in Human-in-the-loop approaches

Methods to audit, measure, and evaluate AI safety:

• Auditing methods and tools

• Benchmarks and case studies

• Standard and best practices

• Explainability, traceability, data and model lineage

• Visual analytics and HCI for understanding/auditing AI for safety aspects 

• Software engineering (design and testing) approaches

Applications

• AI in Safety-Critical Systems

• SafeAI for Peace 

• SafeAI in Healthcare

• SafeAI in Cybersecurity

• SafeAI in Education

• SafeAI in Manufacturing

• SafeAI in Autonomous systems

• Safe resource allocation

EXTENDED DEADLINE

• • Paper submission deadline: 30 May 2025 13 June 2025

  • Reviewing period: 1-16 June 2025 1-17 June 2025

  • Paper author notification:  18 June 2025

  • Camera ready deadline: 1 July 2025

  • Program and papers available online: 10 July 2025

Papers should be submitted on the Safe AI Microsoft CMT submission site Microsoft CMT Conference Link.

Please create first a Microsoft CMT Account here if you do not have one to be able to submit your work. 

It is strongly recommended that you sign-up for Microsoft CMT with an institutional email address.

There will be two paper formatting options: (a) full papers and (b) extended abstracts.

a) Full papers must follow the UAI 20205 main conference style, with same length restrictions: up to 8 pages for the main body + unlimited pages for references and appendices. Please see the UAI 2025 Submission Instructions for more details on how your manuscript should be formatted using the Latex template.

b) Extended abstracts must have 2 pages comprising already published work or an early work in progress.

Lastly, at least one author must attend the event in person.

Single-blind peer-reviewing by the Program Committee.

There will be no formal proceedings. Therefore, authors are allowed to submit their work accepted to the SafeAI workshop elsewhere .

A journal special issue is intended.

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