individual
AI safety
| BlueDot Impact | $3,350 | 100% | 1 |
| Date ↓ | Funder | Via | Amount | Cause | Source | Purpose |
|---|---|---|---|---|---|---|
| Sep 2026 | BlueDot Impact | $3,350 | AI safety | bluedot | We test whether LLMs give value-consistent answers across languages. We find that language shifts responses more than stated identity, skews toward US views, and changes harmfulness judgments. |