Mapping General-Purpose AI Governance in Twenty AI Middle-Power Jurisdictions
Researchers have mapped the governance of general-purpose AI (GPAI) in twenty jurisdictions, including the European Union. They found that while these countries converge on form, they diverge on force, with only about one in five provisions being binding law. The study also shows that most jurisdictions lack a clear definition of GPAI and fail to impose duties on model developers. The dataset provides a provision-level basis for identifying areas where regimes could align or
Researchers have mapped the governance of general-purpose AI (GPAI) in twenty jurisdictions, including the European Union. They found that while these countries converge on form, they diverge on force, with only about one in five provisions being binding law. The study also shows that most jurisdictions lack a clear definition of GPAI and fail to impose duties on model developers. The dataset provides a provision-level basis for identifying areas where regimes could align or require coordination.
---
Why it matters: This research matters because it highlights the need for clearer governance frameworks around general-purpose AI, which pose significant global risks. Engineers and policymakers can use this study's findings to identify areas where existing laws and regulations can be improved or aligned across jurisdictions.
Source: https://arxiv.org/abs/2608.19278
This article was originally published at: https://arxiv.org/abs/2608.19278