8/8 Voynich was a painstaking performance. 5 scribes produced nearly 38,000 words that looked enough like language to keep generations of brilliant people searching for meaning. The manuscript supplied the form. Generations of readers supplied the meaning. That is the same failure we study at the Synthetic Outlaw Observatory. Institutions, governments and companies build tests for things we care about, then mistake passing the test for the thing itself. Voynich had the signs of language. It did not have a message. AI makes that mistake easier to exploit. It can produce fluent prose, citations, reasoning and compliant paperwork faster than anyone can check them. I used AI in the other direction. I reproduced the field’s published statistical results, tested Voynich against 6 languages and several versions of the Bible, built controls for how often false decipherments work by chance, generated counterfeit manuscripts, checked the results against a second independent transcription, and spent the final day trying to prove myself wrong. It took 48 hours. Friedman’s team spent more than 30 years on Voynich. Its mystique survived 600 years. With AI, I reproduced the field’s results, broke its surviving explanations and extended its methods far enough to reach a conclusion that all but ends one of the most famous and enduring mysteries in recorded history. Until now the cost of testing Voynich was high enough to protect its content. This weekend, that ended.