0
I’m curious whether GPT, Gemini, and Claude can make candidate affidavits easier to compare without making them seem simpler than they are. For example, could they pull out declared assets, liabilities, and pending cases from two affidavits while keeping the figures tied to the right candidate and election?
How would you check that a summary hasn’t confused a candidate’s own declaration with a proven fact, or missed a detail in the original filing? If you’ve tried this with Indian election records, which model handled the documents most carefully?
