Experts Warn of AI-Driven Disinformation Threat to UK and US Elections

In a congressional hearing, CEO of OpenAI, Sam Altman, warned that AI models could manipulate humans and pose a significant area of concern. Next year’s UK and US elections in particular could be targeted by a wave of AI-powered disinformation, as generated fake news stories, images, video, and deepfake videos go viral with the help of swarms of propaganda bots. Breakthroughs in generative AI such as ChatGPT and Midjourney raise the prospect of interactive election interference at scale, with AI trained to repeat talking points targeted at specific audiences. Concerns over the technology have led the UK’s prime minister to vow to lead on limiting the dangers of AI.