AI governance should be inclusive, transparent, multidisciplinary, multilateral and multi-stakeholder, covering anticipation, protection, impact monitoring, enforcement and redress. Harms caused through AI should be investigated and redressed through strong enforcement and remediation mechanisms, including those of companies, supported by auditability and traceability. Public authorities should transparently self-assess existing and proposed AI, including whether AI is appropriate and whether it would breach human rights obligations, and prohibit it if so. Rights-sensitive public uses (law enforcement, welfare, employment, media, health care, the judiciary) need monitoring by oversight authorities, including data protection authorities; judicial capacity should be strengthened and judicial use of AI safeguarded by human oversight and judicial independence.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.