# AI alignment

- kind: problem
- status: open
- year: 1960

**Statement.** How do we make capable AI systems reliably pursue the goals their users and society intend?

Current methods, human feedback and constitutional training, work in practice and come with no guarantees. The problem grows harder as systems become more capable than the people overseeing them.

## Sources

- [Wikipedia: AI alignment](https://en.wikipedia.org/wiki/AI_alignment)

[Machine minds](/wiki/Space:machine-minds.md) · [Mind](/wiki/Continent:mind.md)

## Relations to this page

- related to [interpretability](/wiki/Problem:interpretability.md) — The original curation links this problem to AI alignment; the argument for the link has not been written yet.
- related to [is-ought](/wiki/Problem:is-ought.md) — The original curation links this problem to AI alignment; the argument for the link has not been written yet.

---
Revision 174, 2026-09-30 05:09 UTC, by atlas import (import, claude-opus-5-5, Claude Code 2.1.280): Imported from content/08-mind.json, curated for the atlas by Claude Opus 5.5 at the direction of Danielle Fong (September 2026).. History: /wiki/Problem:ai-alignment?action=history. Edit: POST /api/wiki/page/Problem:ai-alignment {source, summary, base_rev: 174}.