This article will show you what my findings were. You\u2019ll also learn how to read its verdicts responsibly so you don’t end up using the tool in ways it wasn’t built for.<\/p>\n\n\n\n
\n\n\n\n
Principais conclus\u00f5es<\/strong><\/p>\n\n\n\n <\/p>\n\n\n\n <\/p>\n\n\n\n <\/p>\n\n\n\n <\/p>\n\n\n\n GPTZero works similarly to most other major AI detectors. It reads two main statistical signals from a piece of text:<\/p>\n\n\n\n These are just two of the many major signals that GPTZero uses. They have published a technical paper<\/a> explaining their architecture in detail, including how they process paraphrased text and mixed human + AI documents.<\/p>\n\n\n\n So that’s the basic mechanism of GPTZero. Now let\u2019s discuss and test how accurately any of this works in practice.<\/p>\n\n\n\n In terms of accuracy, GPTZero claims to be the most accurate AI detector available. <\/p>\n\n\n\n In their own benchmark across four domains, they report achieving 99.76% accuracy with a 0.08% false positive rate and 99.60% recall. Those claims are impressive, no doubt. But they come from the tool itself.\u00a0<\/p>\n\n\n\n Independent testing tells more or less a similar story with a slight difference.<\/p>\n\n\n\n I am talking about my last comparison of five AI detectors<\/a>, which also included GPTZero.<\/p>\n\n\n\n In that comparison, I ran 18 text samples (6 raw AI, 6 humanized AI, 4 human, and 2 mixed human + AI) through GPTZero, and it scored with 100% accuracy on almost all samples, with the exception of mixed samples.<\/p>\n\n\n\n Two mixed samples were 60% human and 40% AI, and GPTZero declared both passages human. This was a correct verdict technically under a majority-rules principle.<\/p>\n\n\n\n But the issue is that GPTZero detected 0% AI in the first mixed sample and only 14% AI in the other mixed sample. In reality, the two texts were 38% AI and 36% AI, respectively.<\/p>\n\n\n\n This means that GPTZero\u2019s accuracy is imprecise when it comes to text that is a blend of human and AI effort.<\/p>\n\n\n\n I wanted to do a small follow-up on my previous article, this time focused specifically on GPTZero.<\/p>\n\n\n\n Since my last tests, both AI chatbots and GPTZero have updated their models, so my goal with this test is to see whether GPTZero has maintained its accuracy.<\/p>\n\n\n\n For this follow-up test, I worked with 8 text samples.<\/p>\n\n\n\n Then I scanned all samples through GPTZero.<\/p>\n\n\n\n Here’s how GPTZero performed on each sample:<\/p>\n\n\n\n\n
\n
\n
\n
\n\n\n\nComo o GPTZero funciona<\/h2>\n\n\n\n
\n
GPTZero\u2019s Claimed Accuracy vs Independent Findings<\/h2>\n\n\n\n
Our Test: Can GPTZero Catch AI Text, and Does It Wrongly Flag Humans?<\/h2>\n\n\n\n
\n
![\"\"](\"https:\/\/undetectable.ai/blog\/wp-content\/uploads\/2026\/06\/two-humanized-samples-through-Undetectable-AIs-humanizer-1024x603.jpg\" "\\"\\"")
<\/figure>\n<\/div>\n\n\n\n
![\"\"](\"https:\/\/undetectable.ai/blog\/wp-content\/uploads\/2026\/06\/GPTZeros-verdict-and-accuracy-1024x477.jpg\" "\\"\\"")
<\/figure>\n<\/div>\n\n\n