Should we assume language models released by Twitter have injected content praising Hunter Biden?
But we should have better tools to test for biases/toxicity. Perspective API is great tool for toxicity detection. But I'm not aware of any "propoganda" detection tool.