Publicly report advanced AI systems' capabilities, limitations and domains of appropriate and inappropriate use, supporting sufficient transparency and so increasing accountability. Code of Conduct detail: publish transparency reports with meaningful information for all significant new releases; keep reports, instructions for use and technical documentation up to date, covering the evaluations run for safety, security, societal and human-rights risks, capabilities and significant performance limits that bear on appropriate use, discussion of effects and risks such as harmful bias, discrimination, privacy threats and fairness, and red-teaming results used to judge fitness beyond development; make the information clear enough for deployers and users to interpret outputs and use the system appropriately; and support reporting with robust documentation.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.