Research & articles
Essays, research notes and evidence reconstructions about security, automation and verifiable operations.
Featured · Essay ·
When Civilization Can No Longer Understand Itself
AI does not need to rebel to become dangerous. The deeper risk begins when humans can still operate civilization but can no longer independently understand, challenge, reconstruct, or recover it.
Latest
Research note ·
How to read a public exposure snapshot
Why a completed check is not a security grade, how request attempts differ from collected responses, and where the evidence stops.