US Government Accountability Office issues AI accountability framework
The US Government Accountability Office has issued an artificial intelligence (AI) accountability framework designed to help federal agencies and other entities ensure accountability and responsible use of AI in government programmes and processes. The framework is organised around four key issues – governance, data, performance, and monitoring – and it highlights practices for implementation by federal agencies and other entities that are considering, selecting, and implementing AI systems. When it comes to governance issues, the framework highlights the need to promote accountability by establishing processes to manage, operate, and oversee implementation of AI systems. For instance, entities should define clear goals, roles, and responsibilities, demonstrate values and principles to foster trust, develop a competent workforce, engage stakeholders with different perspectives to mitigate risks, and implement an AI-specific risk management plan. With relation to data, the framework includes a recommendation for entities to ensure quality, reliability, and representativeness of data sources and processing. Entities are also advised to ensure reliability and relevance over time for the AI systems they use, as well as to define metrics and assess the performance of AI systems.