AI authentication tools are built without adequate journalist input, new report finds

According to a new report from Princeton's Center for Information Technology Policy, tools being built to help journalists authenticate media in the age of AI are largely being developed without adequate journalist input. The challenge is pressing: AI can now convincingly generate photographs, videos, documents, and audio. In response, initiatives like C2PA—launched in twenty twenty-one by Adobe, Microsoft, the BBC, and others—embed cryptographic signatures in media files to record their complete history. But newsrooms report that their publishing software routinely strips these signatures, and social media platforms delete them entirely. Similar problems plague AI detection tools, which often provide confidence ratings like 'eighty percent likely AI-generated' rather than a clear yes or no. The report's authors call for technologists and journalists to develop these tools together from the start, rather than one side building for the other.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton