llama2-license-700m-dau-restriction

IN premise — entries/2026/06/21/wiki-LLaMA-chunk-2.md

Created 2026-06-21T09:50:09+00:00

Llama 2's license blocks entities with >700 million daily active users and prohibits using outputs to improve other LLMs

Summary

Meta's Llama 2 license draws two hard lines: very large companies (above roughly 700 million daily users) simply can't use the model, and no one is allowed to feed its generated text back into training another language model. Any system built on Llama 2 must treat these as legal constraints, not just policy preferences, and cannot assume it can freely deploy at scale or use the model as a training-data generator for other AI systems.

Dependents

These beliefs depend on this one:

IN open-weight-models-face-unresolved-definitional-tensions — The "open" AI ecosystem faces unresolved tensions: Llama's license restricts large platforms and prohibits competitive training use, the FSF classified it as nonfree software, and the OSAID requires training data disclosure that most "open" models do not provide.