Several resources provide overarching guideposts for OFAI research, convenings and activities, including the White House October 2023 Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence, NIST AI Risk Management Framework and US Artificial Intelligence Safety Institute, Research Data Alliance (RDA) Artificial Intelligence and Data Visitation Working Group, Omidyar Network’s position on generative AI, Noble Reach Foundation’s talent programs and academic partnerships, and multi-philanthropies initiative to ensure AI advances the public interest.

Each of these documents outlines a range of important goals and principles. These five representative pillars below from the philanthropies document can serve as guiding principles and overarching vision for OFAI:

1. Ensure that AI protects democracy and the rights and freedoms of all people;
2. Leverage AI to innovate in the public interest and deliver breakthroughs to improve quality of life for people around the world;
3. Empower workers to thrive amid AI-driven changes across sectors and industries;
4. Improve transparency, interpretability, and accountability for AI models, companies, and deployers; and
5. Support the development of international AI rules and norms.

Additionally, OFAI will explore the role of openness for safety, innovation, competitiveness and inclusion.

OFAI will adopt a multi-threaded approach for achieving this vision with the following working groups:

1. Research
2. Technical prototypes
3. Coordinating of research and policy recommendations
4. Community engagement and relationship building
5. Talent for service

OFAI deliverables in 2024 will include the following:

• Position papers and convenings that are inclusive and diverse in representation
• Recommendations for AI policy development in the US
• Development of AI technical prototypes
• Original research, particularly toward the Openness in AI technical framework
• Infrastructure for identifying and training early career talent for AI public service roles

The convenings will include both those aimed directly at policy development and also those that bring together individuals or groups who are not yet well represented in the AI landscape. For example, CMU’s Dietrich College of Humanities and Social Sciences has held initial dialogue with the American Indian Higher Education Consortium to redesign their cyberinfrastructure and culturally responsible implementation of AI and technology across their 38-college system.