Assessment results can be central to using and developing AI responsibly, so they should be analysed and used in both technical and management decisions. Technically, the analysis can improve the product or service, add safeguards against unintended use or foreseeable malicious use, strengthen the system's safety and robustness, and identify the kinds of measures to secure benefits and prevent harms described for benefits and harms in clause 6; results should also be compared with the thresholds set under 5.7, with a corrective action plan where a threshold is not met. At management level the analysis feeds the organization's risk management with information on the foreseeable impacts of AI, which can include looking at single assessments or at several together to spot trends that affect the organization's risks and objectives as a whole.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.