The US Constitution considered one of the most vital documents in the United States, was recently identified by the software as AI-generated in an alarming test.
As ChatGPT’s popularity increases, it is becoming one of the most widely used applications on the internet. With many students turning to it for university essays, school work, and beyond, AI detection software is being developed rapidly to detect those relying on ChatGPT too heavily.
Despite the existence of AI detection software that promises perfect accuracy, it appears to be flawed in its current state, as a variety of authentic documents are being mistakenly recognized as AI-generated works.
Despite the worrying implications, AI detection tools are incorrectly flagging the United States Constitution as AI-generated content, as pointed out by Twitter user Gaut. This highlights a test of these text detectors which provides an undue result and emphasizes that further development within this field is necessary.
Inspection of the IS constitution reveals that 92.26% of it is generated by AI, which far surpasses the acceptable level of 30%. Such a result exposes organizations’ lack of preparedness to address issues related to artificial intelligence.
AI-detection services, like ZeroGPT or TurnItIn, are created based on existing plagiarism checkers. Any text that is identical to the American Constitution – for instance, if it were posted verbatim – would be seen as an act of stealing and thus would be picked up by these systems.
Despite being in circulation for over 230 years, an intriguing 8% of the US Constitution has not been detected as generative work.
Students may already experience issues with some plagiarism checkers over-interpreting citations and quotes. On top of that, the addition of ChatGPT detection could cause a considerable amount of difficulty for those completing essay work. Universities, colleges, and schools have thus far failed to regulate AI written essays and tools sufficiently,
AI-generated works are unreliable enough to be trusted unless they can be easily detected for vetting content online and in the classroom. With accurate detection being a requirement, the current inaccuracy of AI-generated works makes them unusable.
The fact that the US Constitution was flagged as AI-generated content by the ChatGPT detector provides a fascinating glimpse into the capabilities of modern language models while highlighting the need for ongoing research and development to ensure that we can continue to distinguish between human and machine-generated content in the years to come.