Claude vs GPT-5
Identical axes, shared 168-hour window, identical sample disclosure.
Claude
model · anthropic
- Deepfakes & deception7%
- Privacy & data6%
- Existential risk0%
- Jobs & economy0%
- Fear
- 17
- Trust
- 11
- Polarisation
- 79
n = 77 · effective 69
GPT-5
model · openai
8 observed mentions in this window. The coverage is visible; a sentiment ranking is published once the effective sample reaches 20.
n = 8 · effective 8