According to TechCrunch, three decades of attempts to control cybersecurity software exports have all failed. From PGP encryption in the nineteen-nineties to today, regulators have struggled to contain tools that inherently resist borders—source code travels at light speed, and knowledge can't be unlearned. Now the question repeats: Anthropic's new Mythos cybersecurity model faces the same export-control scrutiny. But history suggests it won't stick. Digital security tools are designed to spread, hard to monopolize, and simple to reimplement elsewhere. Restricting one version doesn't prevent researchers from building another. The piece raises a fundamental tension in tech policy: governments assume they can control the flow of software. Three decades of evidence says otherwise.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton