Digital colonialism

Posted: 2nd August 2026

The first was the story about the massive scale with which AI companies are burning the antique books they’ve used to train LLMs.

The second was how a propaganda firm was paid $46.5 million by Israel to stack the internet with hundreds of baseless smears, such as posts asserting that any journalist killed in Gaza was a terrorist. But here’s the rub: the intended audience for these posts are not humans; the goal is actually to influence chatbots like Claude and ChatGPT and Grok!

As Nvidia chip founder Jensen Huang recently explained, the future of AI depends on continuous knowledge sharing between AI models. (Some people call this “LLM poisoning.”) The idea is that they will then incorporate the “poisoned” framing into all the material they generate - both as part of phony AI websites and as answers to queries - so that these lies ultimately dominate our collective information space.

When these two stories are read together one after another, a deeply disturbing pattern emerges. A rush is on to destroy the legacy of human culture and replace it with a fiction-filled, privatized digital twin that has been bought, paid for, and - because of its digital nature - is utterly controllable. This tactic is at the heart of colonialism’s playbook.

—