Actuaries Flag Agentic AI, Climate Model Gaps as Risks
Actuarial publications are flagging agentic AI, electric motorcycles, and hydrogen fuel as key emerging risks for 2026 that will require new data sources and risk models. Separately, FTI Consulting warned that current climate risk models may be underestimating financial exposure by up to a factor of four.
- A joint report from the Institute and Faculty of Actuaries (IFoA) and the University of Exeter warns that many climate scenarios used by financial firms are "far too benign" because they exclude severe impacts like climate tipping points, sea-level rise, and second-order effects like civil unrest. - One key issue with agentic AI is the risk of "group think"; if multiple insurers adopt similar frontier models, it could introduce new systemic risks to the market as the models may respond in the same way to market shocks. - The risk profile for electric motorcycles is altered by their rapid acceleration and near-silent operation, which makes them more likely to be involved in accidents with pedestrians who may not hear them approaching. - Hydrogen fuel presents unique physical risks, including being highly prone to leaks and burning with a nearly invisible flame, which complicates detection and firefighting. - A follow-up analysis warns that global temperatures are accelerating faster than predicted due to the loss of a hidden "sunshade" effect from air pollution, which had been offsetting warming by about 0.5°C. - For insurers, agentic AI creates potential exposures in errors and omissions, cyber liability, and product liability related to the actions and decisions of the AI agents. - The global market for insuring hydrogen projects is projected to surpass $3 billion in premiums by 2030, driven by a sevenfold increase in planned projects between 2021 and 2024. - A significant risk in hydrogen infrastructure is "hydrogen embrittlement," a process that can weaken and cause cracking in steel pipelines and storage components.
