As a first part of CIPL’s EU AI Act Implementation Project and in conjunction with ongoing research on the responsible and accountable development and deployment of artificial intelligence systems, CIPL has identified Article 4 AI literacy best practices and recommendations for practitioners.
AI Literacy Best Practices for Whom?
These best practices can be adapted and evolved depending on the size of an organisation or the extent to which it relies on AI systems. For organisations that are in nascent stages of AI adoption, this white paper provides best practices for AI literacy to support responsible and trustworthy AI initiatives. Equally, organisations that have been developing and deploying AI systems over a longer period with mature AI programs, including AI literacy programs with resources and tried and tested strategies, can profit from benchmarking against other AI literacy best practices.
CIPL’s Article 4, AI Act AI literacy best practices and recommendations include:
- Ensure organisational leadership signals support for and prioritization of AI literacy programming.
- Leverage and integrate AI literacy obligations in existing compliance programmes and business goals.
- Consider an organisation’s workforce, contractors, third party vendors, clients/ customers, users, and individuals affected by the organisation’s use of AI systems when developing AI literacy programs.
- Make AI literacy an organisation-wide, shared effort to promote existing knowledge, foster new learning, and identify best practices.
- Promote a common, organisation-wide AI taxonomy that includes an inventory of fully assessed use cases.
- Take a contextual approach to AI literacy training that considers role, expertise, and diverse learning preferences.
- Look beyond training by creating opportunities for stakeholders to put the knowledge into practice with support.
- Establish a continuous feedback loop for AI literacy programming and adjust programming and training based on feedback and ongoing developments.
