While governments typically procure and use models developed by the private sector (which is covered in the “Procurement & Usage in AI deployment” section below), there is some evidence of governments influencing the development of models through their use. Canada’s Directive on Automated Decision-Making requires federal institutions to conduct Algorithmic Impact Assessment, thereby influencing the development of models to minimise the risks assessed in relevant tools. A more direct example is DARPA’s XAI Program, where the need for explainability in AI models used in operational military contexts (for legal and ethical compliance) resulted in the Explainable AI Toolkit, which contains a variety of tools and resources to help users, developers, and researchers understand outputs from machine learning models beyond military applications. In Chile, public-sector AI projects must follow public tendering AI standards set by ChileCompra, the government’s procurement agency under the Ministry of Finance, which may impact the long term development of such systems.
Government Interventions in AI
Please click on any quadrant to read the taxonomy.
< Back to main page