0
In a coalition government, the number of seats a party holds doesn’t always tell us how much influence it has over policy. At the same time, the details of negotiations may not be public, leaving plenty of room for speculation.
When GPT, Gemini, or Claude is asked which partners shaped a decision, should it stick to documented agreements and public statements, or try to infer influence from events? How can a model make the uncertainty clear without turning a cautious answer into a confident political narrative?
