The New York Times Tested Every AI Fake Detector and the Results Are Terrifying

Can you tell if a photo is real or made by AI? Neither can most of the tools designed to figure that out.

The New York Times just published a massive investigation where they tested more than a dozen AI detection tools with over 1,000 images and videos. These are the tools that social media platforms, fact-checkers, and regular people use to figure out if something they see online is real.

The results are honestly scary. While some tools did okay at catching certain types of AI content, none of them were accurate enough to give users complete confidence. Some tools flagged real photos as AI-generated. Others said obviously fake AI images were totally real.

They tested tools from major companies including Google's Gemini Pro, OpenAI's ChatGPT, Anthropic's Claude, and Elon Musk's SuperGrok, along with specialized detectors like Hive Detect, Sensity, and Reality Defender.

The bottom line? Right now, there is no single tool you can trust to definitively tell you if something is AI-generated. The NYT recommends using multiple detectors and backing up their results with other research.

This matters because AI-generated images and videos are flooding social media at an unprecedented rate, and most people have no reliable way to tell what's real anymore. As reported by The New York Times.

Source: The New York Times