New York-based Pangram has raised $9 million to combat the growing tide of AI-generated content on the internet. Their latest model, Pangram 4, boasts a 99% accuracy rate in detecting AI-assisted writing and mixed human-AI content.

The startup's co-founders, Stanford AI graduates Max Spero and Bradley Emi, launched Pangram after the release of ChatGPT sparked a flood of bot-generated content. Their technology is designed to help users discern whether text was written entirely by humans or with some AI assistance.

Pangram’s detection system learns from millions of human documents, identifying stylistic differences made consistently by AI. It can automatically label posts on platforms like X, LinkedIn and Substack, providing a feed health score that breaks down the percentage of content created by humans versus machines.

As AI usage becomes more commonplace, so too do the potential consequences of accepting AI-generated content at face value. Institutions are starting to introduce policies to address this issue, while competitors like Winston AI and Copyleaks are also entering the fray. Pangram’s technology could play a significant role in resisting the flood of AI content.