NYT Adopts AI Detection for Photos and Videos

- The New York Times is a co-founder of the Content Authenticity Initiative (CAI), launched in 2019 with Adobe and Twitter, to develop an open standard for digital content provenance. This initiative aims to combat misinformation by creating a way to verify the history and origin of media. - The technical standards for content provenance are being developed by the Coalition for Content Provenance and Authenticity (C2PA), an organization that includes Adobe, Microsoft, Intel, and the BBC. Their "Content Credentials" act like a nutrition label for digital files, showing details about the content's origin and any subsequent edits. - This move is part of a larger legal battle; in December 2023, The New York Times sued OpenAI and Microsoft for copyright infringement. The lawsuit alleges that millions of its articles were illegally used to train large language models like ChatGPT, which now compete with the newspaper. - The lawsuit demands that OpenAI and Microsoft be held responsible for billions of dollars in damages and calls for the destruction of any AI models and training data that use The Times' copyrighted material. - While AI detection tools are being adopted, tests have shown they are not yet accurate enough to provide complete confidence. They perform better at identifying real images than AI-generated fakes, creating a technological "arms race" as AI generators constantly improve. - The rise of sophisticated deepfakes has created an "authenticity crisis," blurring the line between documentation and fiction and eroding public trust in visual media. This has put pressure on newsrooms to be more transparent about any digital alterations beyond basic cropping or color correction. - This is not The Times' first use of AI; in 2018, it began using Google's AI technology to digitize and analyze its vast physical photo archive, which contains 5 to 7 million photos dating back to the 1870s.