New York-based AI detection startup Pangram is on a mission to combat the AI slop infestation spreading across the internet, and it just raised $9 million on a bet that demand for tools that distinguish human-generated content from AI-generated text will only grow. Pangram's fundraise, led by Menlo Ventures with participation from Haystack, Script Capital, and Cadenza, comes as the startup also launches its next-generation AI text detection model, Pangram 4, and an AI image detection model, Pangram Image. Pangram says the new text detection model is over 99% accurate at finding AI-assisted writing and mixed human-AI content. Plus, it can more easily detect AI humanizer programs.

Is this something that I'm going to have to look out for hallucinations and jump in skeptically? Or is this something that I trust was well-researched from an actual journalist? Pangram's AI detection system is essentially a large machine learning model that was trained on tens of millions of known human documents. The startup then created a synthetic mirror for each document, replicating the topic, length, and tone of voice, but written by a frontier LLM. Spiro said, Our model is learning the stylistic differences and the choices that AI makes consistently, and is able to use that to learn what makes something AI-generated with high confidence.