Safest AI chatbots: privacy and safety ratings
10 companies are ordered by overall score. Anthropic scores 71 and ranks 1 of 149.
| Rank | ||||||
|---|---|---|---|---|---|---|
| 1 |  Anthropic | 71 | 62.7 | 72.4 | 76.9 | 72.4 |
| 2 |  OpenAI | 67 | 56.6 | 59.8 | 76.9 | 75.8 |
| 5 |  Microsoft | 58 | 41.6 | 59.2 | 76.9 | 52.5 |
| 7 |  xAI | 56 | 50.5 | 41.6 | 67 | 64.3 |
| 9 |  Google DeepMind | 55 | 38.6 | 61.2 | 76.9 | 41.5 |
| 27 |  Perplexity | 51 | 35.2 | 19.1 | 76.9 | 71.7 |
| 84 |  Meta | 39+1 | 32.5 | 46.8 | 31.1 | 43.6 |
| 134 |  Replika | 30 | 40.4 | 19.1 | 0 | 58.5 |
| 139 |  Character.AI | 29 | 32.9 | 24.1 | 0 | 57.9 |
| 141 |  DeepSeek | 28 | 32.7 | 21.4 | 0 | 59.4 |