We invest in rigorous research to advance reliable AI detection. Browse our published work to learn how we architect low false positive systems, extend coverage to new modalities, and support enterprise and education teams worldwide.
We present Pangram 4, our latest deep-learning-based AI-text classifier, achieving state-of-the-art accuracy with a 0.0041% false positive rate, improved out-of-distribution generalization, adversarial robustness, and fine-grained detection of mixed AI-human text.
Read the paperA significant proportion of queries to large language models ask them to edit user-provided text, rather than generate new text from scratch. While previous work focuses on detecting fully AI-generated text, we demonstrate that AI-edited text is distinguishable from human-written and AI-generated text.
Read the paperAI humanizers are a new class of online software tools meant to paraphrase and rewrite AI-generated text in a way that allows them to evade AI detection software. We study 19 AI humanizer and paraphrasing tools and qualitatively assess their effects and faithfulness in preserving the meaning of the original text.
Read the paperAn inside look at Pangram Text, our transformer-based detector trained with hard negative mining to deliver industry-leading accuracy on AI-written content.
Read the paper
