AI-driven disinformation floods social media before US midterms as platforms pull back protections

Social media feeds are being inundated with an unprecedented wave of AI-generated disinformation in the run-up to the November US midterm elections, experts warn, with many deceptive posts likely slipping through the cracks. Major platforms, including Meta, have scaled back safeguards designed to shield voters from election-meddling lies, leaving tens of millions of Americans vulnerable to manipulation as they prepare to cast their ballots. The rollback of content moderation and fact-checking measures has created fertile ground for automated propaganda, deepfakes, and false narratives aimed at swaying public opinion and undermining trust in the electoral process. Researchers caution that the sheer volume and sophistication of AI-produced content make it harder than ever for both users and platforms to distinguish fact from fiction. With election day approaching, the stakes are high: unchecked disinformation could influence outcomes in key races and further erode confidence in democratic institutions. The takeaway: as artificial intelligence supercharges the spread of falsehoods, the retreat of platform accountability leaves voters more exposed than ever.