Constitutional AI
ai_safety
Overview
Developed byAnthropic
Use casetraining harmless AI using a set of written principles instead of human labels
Also see
Alternative to
Knowledge graph stats
Claims8
Avg confidence95%
Avg freshness99%
Last updatedUpdated yesterday
Trust distribution
100% unverified
Governance
Not assessed
Constitutional AI
concept
Anthropic alignment technique training AI with a set of principles rather than human feedback alone
Compare with...alternative to
| Value | Trust | Confidence | Freshness | Sources |
|---|---|---|---|---|
| RLHF | ○Unverified | High | Fresh | 1 |
used by
| Value | Trust | Confidence | Freshness | Sources |
|---|---|---|---|---|
| Claude | ○Unverified | High | Fresh | 1 |
primary use case
| Value | Trust | Confidence | Freshness | Sources |
|---|---|---|---|---|
| training harmless AI using a set of written principles instead of human labels | ○Unverified | High | Fresh | 1 |
first released
| Value | Trust | Confidence | Freshness | Sources |
|---|---|---|---|---|
| 2022 | ○Unverified | High | Fresh | 1 |
developed by
| Value | Trust | Confidence | Freshness | Sources |
|---|---|---|---|---|
| Anthropic | ○Unverified | High | Fresh | 1 |
implemented by
| Value | Trust | Confidence | Freshness | Sources |
|---|---|---|---|---|
| Anthropic | ○Unverified | High | Fresh | 1 |
| Claude | ○Unverified | High | Fresh | 1 |
governed by
| Value | Trust | Confidence | Freshness | Sources |
|---|---|---|---|---|
| AI Constitution | ○Unverified | Moderate | Fresh | 1 |