IRGC’s strategy was proactive. Our intention was to help decision-makers, particularly policymakers, anticipate and understand emerging and systemic risks, as well as risk governance options, before they become urgent policy priorities. We identified issues with risk governance deficits and developed recommendations for improving their assessment, evaluation, communication and management.
For example, in 2018 we drew attention to risk that machine learning algorithms used in AI systems would pose when used for automating decisions (decision-making algorithms), without human control. In 2019, we drew attention to deepfakes and risks posed to misinformation and manipulation (forged authenticity). In 2023, the advent of generative AI used in large-language modelling (LLMs) such as GPT confirms the risk, with broad concerns that generative AI could pose to society, and truth and trust models.
IRGC’s process
- Identifying potential risk issues at the earliest possible stage
- Understanding the issue and the associated risks, as well as the institutions, risk governance structures and processes that are currently in place for assessing and managing the risks
- Identifying governance gaps that hinder the efficacy of the existing risk governance structures and processes
- Making recommendations for overcoming these gaps.
IRGC’s values
Within its mission to improve global risk governance, IRGC was:
- Open: project outcome is shared freely
- Accountable: science-based project work and recommendations are scrutinised via peer review before publication
- Collaborative: this lies at the heart of our approach and integrated working is vital
- Independent: this allows us to freely choose the subjects on which to focus, select experts and partner organisations with whom to collaborate, and design the appropriate governance recommendations to deal with the risks we address.