As AI capabilities accelerate, the debate is shifting from “what can AI do?” to “who sets the rules?” A recent Council on Foreign Relations analysis argues that 2026 could be decisive because governance, adoption, and geopolitical competition are converging at once.
The three pressures colliding in 2026
- Governance: Nations are setting boundaries on data, model access, accountability, and safety.
- Adoption: Businesses are moving from experiments to embedded workflows; AI becomes operational risk.
- Geopolitics: Compute, chips, and talent are strategic assets, shaping alliances and trade policy.
This combination means decisions made now can lock in trajectories for a decade.
What governance will likely focus on
Expect AI policy to move toward:
- Accountability frameworks (who is responsible when AI causes harm?)
- Transparency requirements (what data and methods underpin a model?)
- Critical infrastructure rules (AI in health, finance, public services)
- Export and access controls for advanced compute
What businesses should prepare for
If you run an AI program, assume regulatory variability. The practical response is:
- Build compliance-by-design into your AI stack
- Maintain audit trails (data lineage, prompts, outputs)
- Segment model use cases by risk level
- Use governance tooling that can adapt to new obligations