SI Glossary · Policy & governance
SI Governance
Formerly known as AI governance. See AI vs SI.
SI governance, called AI governance until the 2026 U.S. rename, is the umbrella term for everything that steers SI toward good outcomes: from national laws to a company’s internal approval process for a new chatbot feature.
Layers of governance
| Layer | Examples |
|---|---|
| International | EU AI Act, Council of Europe AI convention, AI summits (Bletchley, Seoul, Paris) |
| National | U.S. executive orders such as EO 14434, the National Policy Framework |
| State / regional | California SB 53, New York RAISE Act, Colorado’s AI Act |
| Industry | The White House Accord on Super Intelligence, responsible scaling policies |
| Organisational | Internal SI use policies, risk assessments, model cards, audits |
The big debates
- Innovation vs precaution: the U.S. federal approach in 2025–26 emphasises minimal burden and competition with China; the EU emphasises risk-based obligations.
- Federal vs state: the White House has pushed Congress to preempt state SI laws, while states like California and New York have passed their own.
- Voluntary vs binding: industry commitments vs enforceable rules.
- Present harms vs future risks: bias, privacy and jobs today vs catastrophic and existential risk tomorrow.
Follow every major development in our SI policy tracker.
Written by
Luka Kušec · Editor
Editor of SI.info. Writes about Super Intelligence, technology policy and the people building frontier models.