The Department of Commerce released on Thursday a guidebook to drive its units’ efforts in making public datasets more accessible to generative artificial intelligence tools.
Titled “Generative Artificial Intelligence and Open Data: Guidelines and Best Practices,” the 79-page roadmap offers actionable steps, such as maximizing documentation availability and accessibility and providing human and machine-readable data documentation.
Additionally, the guidebook suggests that its agencies adopt the appropriate use of open-source software and formats. It also urges them to publish structured and extensive data and metadata.
Future Updates to Expect
The guidance is a first version and will be updated periodically, the department said. Future work identified in the guidebook includes focus on exploring digital signatures to protect data integrity and security, particularly for datasets for developing AI models. Metrics or pointers on evaluating AI-readiness of data assets is another area to be looked into.
A working group composed of the department’s AI and data experts was formed by the Commerce Data Governance Board to initiate the drafting of the guidebook in late 2023. The department issued a request for information on the guidance in April and also announced in May four department-wide draft publications on safe AI development.
Register for the Potomac Officers Club’s 2025 Artificial Intelligence Summit on March 20 to hear more about cutting-edge AI innovations from industry and government experts.
