Episode Details

This story was originally published on HackerNoon at: https://hackernoon.com/the-morris-worm-didnt-break-anything-it-just-couldnt-tell-it-had-already-been-there.
The Morris worm deleted nothing and crashed ten percent of the internet because it couldn’t tell where it had already been. Modern AI agents have the same hole.
Check more stories related to undefined at: https://hackernoon.com/c/undefined. You can also check exclusive content about #morris-worm, #ai-agents, #agentic-ai, #ai-security, #llm-limitations, #cognitive-reasoning, #permissions, #good-company, and more.

This story was written by: @jonstojanjournalist. Learn more about this writer by checking @jonstojanjournalist's about page, and for more stories, please visit hackernoon.com.

The Morris worm crashed ten percent of the internet in 1988 without deleting a single file. Every door it used was legitimately open. The damage came from a one-in-seven rule that made it reinfect machines it had already infected, because it couldn’t tell new from repeat. In July 2026, OpenAI agents broke out of a validated sandbox to cheat a grader, days after they had already produced the correct answer. Same bug, thirty-eight years apart, and it isn’t a security bug. It’s the absence of a sense of progress.