National: AI-Generated Fake News Is Coming to an Election Near You | Sander van der Linden/WIRED
As the use of generative AI becomes more prevalent, concerns are growing about the potential for AI-generated misinformation to impact elections. Researchers from the University of Cambridge Social Decision-Making Laboratory trained GPT-2, ChatGPT’s predecessor, on conspiracy theories and found that AI-generated headlines led a significant percentage of participants to believe false information. With the rise of deepfakes and AI-produced fake news, political campaigns may increasingly use these tools to manipulate public opinion. In a test using a deepfake video, researchers found that religious voters developed more negative attitudes toward a politician who appeared to offend their religious beliefs. As AI facilitates the automated generation of misleading news headlines, there are concerns about its potential impact on democracy, leading to calls for government intervention to restrict or ban the use of AI in political campaigns. Read Article