Key facts
- AI-driven disinformation is flooding social media ahead of US midterm elections.
- Platforms are increasingly relying on AI for content moderation, allowing more falsehoods to spread.
- Campaigns are struggling to counter rapidly spreading online lies, with some evolving into different false claims.
- Social media companies have reduced content moderation efforts, particularly for non-English content.
- AI-generated content, like deepfake ads, is being used in political campaigns.
AI-driven disinformation is proliferating on social media platforms ahead of the US midterm elections, with experts warning that content moderation efforts are insufficient to combat the spread of falsehoods. Social media companies have reduced their content moderation teams and are increasingly relying on AI, which makes it cheaper and easier to launch sophisticated disinformation campaigns. Operatives can continuously tweak posts to bypass moderation rules, leading to lies that spread rapidly and evolve as they propagate. This poses a significant challenge for campaigns, which find it difficult to counter fabrications once they gain traction. Platforms like Meta and X have not responded to requests for comment, while YouTube stated all users are subject to the same guidelines. The issue is particularly pronounced for non-English content, with Spanish-language content facing fewer guardrails. AI-generated content, such as deepfake advertisements, is also being used in political campaigns, raising concerns about its impact on civic discourse.