Safe and secure use of AI in research projects

AI tools increasingly support all stages of research projects. At the same time, their use is constrained by ethical principles, research-integrity standards, and governance (legal and regulatory) requirements. This workshop introduces these constraints, presents the risks of AI use in research, and provides practical strategies for mitigating them across the entire research lifecycle. Both unintentional harms (AI safety) and deliberate threats (AI security) will be addressed. Participants will learn how to use AI tools safely, securely, and in compliance with ethical, integrity, and governance expectations.

Files: slides, zip-archive of GitHub repo (https://github.com/shigapov/safe_ai/) with jupyter-book (https://shigapov.github.io/safe_ai/), and workbook in PDF-format.

Agenda:

INTRODUCTION
PART 1: Theory
1.1 Freedom of research under ethical, integrity, and governance (legal and regulatory) constraints
1.2 AI risks, AI safety, and AI security
1.3 Risk management
PART 2: Practice
2.1 Risk management framework adapted to the trinity of risks (ethical, integrity, and governance)
2.2 Hands-On: Use cases of AI and risk management across the research lifecycle. Plan & design, collect & create, analyze & collaborate, evaluate & archive, share & disseminate, access & reuse.
2.3 AI policies and checklists for research groups and research projects
SUMMARY