Anthropic collaborated with the Institute for Strategic Dialogue and the McCain Institute to test Claude AI's safeguards against election-related misuse across the United States, Israel, and Brazil.