Entity: Pangram
Pangram (Pangram Labs) is an AI detection technology provider specializing in identifying machine-generated text across digital platforms and enterprise environments.
Platform Intelligence & Detection Mechanics
- LinkedIn Content Analysis: Pangram research revealed that approximately 40% of long-form writing on LinkedIn was generated by artificial intelligence models.
- Detection Methodology: Pangram works by measuring statistical “peaks” in language distribution characteristic of Large Language Models (LLMs) resulting from pre-training and reinforcement learning (RLHF).
- Steerability & Fine-Tuning Limits: Text detection tools like Pangram measure whether text passed through an LLM during generation, but can be influenced by model fine-tuning (e.g. models fine-tuned to mimic human variance) or heavy prompt engineering.
Integration in Creator Platforms
- Substack Integration: substack, under CEO chris-best, integrated Pangram’s analysis engine into the Substack app. This enables readers and writers to run voluntary text scans to check for LLM generation signatures.
- Transparency Tooling: Pangram’s metric is used as a neutral “tick of transparency” rather than a enforcement mechanism. It acts as a conversational baseline for creator disclosures alongside “How I made this” process notes.