Both the US and EU approaches to AI governance have their downsides. The scope of the former only imposes a requirement on Federal bodies and relies on voluntary commitments from leading developers. The latter has been criticised for its top- down, prescriptive approach and the potential for uneven implementation across different member states.