Pangram has introduced a new AI text detection model capable of identifying 99.66 percent of AI-generated content while maintaining a very low false positive rate of one per 24,000 documents, according to The Decoder.
The model also shows strong resistance to so-called humanizer tools, which attempt to mask AI-generated writing as human-authored, enhancing its reliability in differentiating authentic text from machine-produced content.
Alongside these improvements, API pricing for the detection service has increased significantly, rising between two- and tenfold. This development is particularly relevant to Japanese markets, where increasing AI integration across FX, crypto, and equities sectors demands robust tools to ensure content authenticity and compliance.