Prioritization of Risks from Artificial Intelligence June 2026 A Delphi Study of 272 International Experts

2026

Open publication workspace · Sign in to read the full PDF.

AI-generated summary

1) This study provides crucial insights into the prioritization of AI risks by surveying 272 international experts, highlighting the most severe threats and identifying key actors responsible for mitigation.
2) * Experts identified dangerous capabilities, weapons & cyberattacks, power centralization, and false information as the most severe AI risks.
* A significant majority of AI risks were judged to have a >10% probability of catastrophic outcomes within the next five years, even with pragmatic mitigations.
* AI users and affected stakeholders are considered most vulnerable, while general-purpose AI developers and governance actors are deemed most responsible for addressing these risks.
3) AI risk, Delphi study, expert elicitation, AI governance, risk prioritization

Check the original publication for accuracy and context.