Pangram Raises $9 Million To Expand AI Text And Image Detection Technology
Pangram has raised $9 million to improve its AI-generated text detection technology and expand into image detection. The round was led by Menlo Ventures, with participation from Haystack, ScOp Venture Capital, Script Capital, and Cadenza.
The company also launched Pangram 4, its latest text detection model, and released a public research preview of an AI image detector.
Pangram 4 is designed to identify fully AI-generated writing, text created jointly by people and AI, and content modified by so-called humanizer tools.
Humanizer programs paraphrase or restructure AI-generated text to make it more difficult for detection software to recognize.
Pangram said the new model maintains a false-positive rate of 0.01%, equivalent to one human-written passage being incorrectly classified as AI-generated out of every 10,000 passages tested.
False positives are particularly important in academic, publishing and workplace environments because an incorrect result could unfairly suggest that an individual used AI.
Pangram also said its image detection model achieved 99.5% accuracy in internal testing. The model remains in research preview, and its performance has not been independently verified.
The company’s systems are used by academic institutions, publishers, journalists, content moderation teams and online platforms.
Pangram plans to use the financing to improve detection accuracy and extend its technology into additional forms of AI-generated content.
KEY QUOTES:
“Fully AI-generated content is everywhere now, and much of it is undisclosed.”
“Pangram exists to make authorship legible for publishers, teachers, journalists and anyone who relies on the truthful and authentic communication of information.”
Max Spero, Co-Founder and CEO of Pangram
“With LLMs, the cost of producing language collapses to zero, and the cost of trusting language goes to infinity.”
Deedy Das, Partner at Menlo Ventures
