Advancing independent research on AI alignment

The Alignment Project is a global fund for independent alignment research created by the UK AI Security Institute.

OpenAI announced a $7.5 million grant to The Alignment Project, outlined its alignment priorities, and Renaissance Philanthropy will support grant administration.

Frontier labs have access to frontier models and significant compute for alignment work, while independent researchers explore conceptual, theoretical, and blue‑sky ideas.

Iterative deployment means increasing capabilities stepwise while strengthening safeguards to surface problems and gather evidence about what works.

A healthy alignment ecosystem relies on independent teams testing diverse assumptions; progress toward AGI may depend on fundamental breakthroughs that change the alignment problem.