Reddit says AI detection cuts exposure to spam and inauthentic content by 20%

Reddit says deploying advanced AI detection models has reduced user exposure to spam and inauthentic content by 20%. The company also says it now revokes nearly 2 million fake upvotes daily, as it outlines protections for the 2026 U.S. elections. Its safety teams combine automated detection with human review to enforce rules and verify automated accounts' origins. AI-generated content remains permitted when it follows community and sitewide rules, but content deliberately misleading people about real events or individuals is prohibited.