AI-Generated News Sparks Trust Crisis: Global Mainstream Media Unite to Establish Human Editorial Certification System

BBC, Reuters, the Associated Press, Agence France-Presse, The New York Times, and The Guardian, along with 36 other global mainstream media outlets, jointly announced the HumanVerified certification system on May 3. Each participating media outlet's news articles will carry a digital watermark indicating the degree of human editorial involvement — from "fully human-reported and edited" to "AI-assisted reporting with human review" to "AI-generated with human review." Readers can see each article's creation mode through browser plugins or on-site indicators.

Reuters Editor-in-Chief Alessandra Galloni, one of the initiators, said: "When AI can generate a seemingly professional news article in 30 seconds, readers have the right to know whether they're reading a journalist's field investigation or an algorithm's text assembly. HumanVerified isn't anti-AI — it demands transparency."

The certification system's backdrop is the explosive growth of AI-generated news content. According to the Reuters Institute's 2029 Digital News Report, approximately 35% of global online news content is entirely AI-generated or AI-assisted, up from just 12% in 2027. More concerning, multiple studies show ordinary readers cannot effectively distinguish AI-generated from human-written news — accuracy rates only slightly exceed random guessing.

MediaTrust Analytics founder Dr. Sarah Chen notes: "HumanVerified addresses information source transparency, but it cannot solve the fact that AI-generated content already exists in massive quantities across the internet. Media outlets and self-media that haven't joined the certification system can still freely publish AI-generated content without any labeling."