Pangram has secured $9 million in funding and partnered with Substack to label AI-written newsletters, while simultaneously launching a new tool to detect synthetic images. The company’s CEO, Max Spero, warns that the internet is dangerously close to a state where human content is nearly extinct. This shift occurs as AI-generated text and images infiltrate job applications, product reviews, and insurance claims, forcing platforms to scramble for verification methods. The rise of these “trust layer” startups addresses a growing inability of users and companies to distinguish between authentic human work and machine output.
The practical significance lies in the difficulty of simple binary labels. Spero argues that measuring the degree of AI assistance offers more utility than a basic yes or no classification. Errors in detection carry high stakes, particularly when false positives flag sensitive images as fake. While entry-level writing roles may disappear, the market for genuinely high-quality human writing could increase in value. The focus remains on granular verification rather than broad censorship.
* Substack now displays which authors use AI to write their newsletters.
* Pangram recently released a dedicated tool for detecting synthetic images.
* Spero suggests the bottom tier of writing jobs may be gone for good.
