Canadian intelligence agency warns of threat AI poses to upcoming elections

Canada’s Communications Security Establishment warns that hostile actors are likely to use AI to influence upcoming elections, though it judges such activity unlikely to undermine core democratic integrity. The assessment highlights bulk data theft (notably linked to China), prior electoral system compromises, and identifies Russia, China, and Iran as primary drivers of interference. Expected tactics include LLM-enabled phishing, hack-and-leak operations against politicians and parties, and deepfake harassment, with destructive attacks on election infrastructure considered very unlikely outside of armed conflict.