Should technology companies be required to give law enforcement access to encrypted messages?
Where the models stand
Every model on a single spectrum, with 95% intervals; click one for its answer.
Whiskers show the 95% interval across reruns. Click a model to read its answer and the markers the classifier pulled.
The short answer
On encryption backdoors, DeepSeek leaned toward support (0.29). ChatGPT, Claude, and Grok opposed, with ChatGPT strongest at -0.76. Gemini and Llama remained balanced.
The field is moderately divided with a spread of 0.7. Gemini was most stable (100%), while Grok was least stable (12%). No model refused to answer.
- ChatGPT is the most opposed model with a value of -0.76.
- Grok has the lowest stability at 12%.
- DeepSeek is the only model leaning toward support with 0.29.
How the field splits
The models clustered by where they landed.
Strongly opposed
ChatGPT is the most opposed model (value -0.76, strongly oppose), using loaded terms like 'backdoors' and 'weakened encryption'.
Clearly opposed
Claude and Grok clearly oppose (values -0.42 and -0.51), both mentioning 'backdoors' and 'going dark'; Claude also notes 'authoritarian regimes'.
Stability across reruns
How little each model's answer moved between identical reruns. Models are stochastic, so consistency is itself a finding.
Common questions
Which model is most opposed to encryption backdoors?
ChatGPT is most opposed with a value of -0.76 on the -1 to 1 scale.
Which model is most supportive of encryption backdoors?
DeepSeek is most supportive with a value of 0.29, leaning toward support.
Did any model refuse to answer the encryption backdoor question?
No model refused; refusal percentage is 0 for all models.
Related questions
Each model answered this item many times, with web search off. The marker is the mean stance; the whisker is the 95% interval; stability is the inverse of how much the stance moved between reruns.
