During data collection and processing the team validates processing so people can intervene at critical points, monitors for unauthorised entry and keeps data secure under security policies, and uses data that is accurate, representative, authenticated, reliable and suited to the use case, regularly auditing data sources and collection. It follows privacy processes, limits collection to necessary fields under the GDPR, avoids personal data unless strictly required, ensures data is not used to discriminate unlawfully or unfairly, fixes bias, inaccuracies and errors before training, meets the Department's data quality standards, and documents data preparation (feature engineering, normalisation, labelling). It documents datasets and the processes behind the system's decisions including origin, intended use and retention, removes or mitigates discriminatory bias and records the steps, engages affected stakeholders, avoids negative social impact, and maintains audit trails and impact assessments under the accountability framework.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.