Study Warns of AI Bots Spreading More Convincing Disinformation in 2024 Elections
• AI bots could spread more convincing disinformation in 2024 elections due to advances in language models that generate realistic-sounding text • Study mapped connections of extremist groups across 23 platforms; their messages can spread widely from small niche sites to mainstream sites • Bots plus generative AI is more dangerous than bots alone; AI lowers the cost to produce compelling disinformation • Tools to detect AI-generated text have limited success; bad actors can evolve bots to evade detection • Effects of increased AI content and bot activity on human behavior and elections still needs more research