AI technology faces scrutiny over unreliable detection and copyright issues

The rise of artificial intelligence is creating significant legal and ethical challenges regarding content authenticity and intellectual property. AI detection tools used by schools, publishers, and social media platforms are facing scrutiny due to their lack of reliability. These tools, such as GPTZero and Turnitin, use statistical probabilities to estimate whether a text was machine-generated, which can lead to false accusations.

Reports indicate that these detectors often produce false positives, harming reputations and careers. For example, a student at Yale was suspended based on an AI detection analysis, and an author lost a $2 million contract after being falsely accused. Furthermore, research from Stanford suggests these tools may harbor biases, disproportionately flagging the writing of non-native English speakers.

In parallel, the unauthorized use of copyrighted material to train or feed AI systems remains a major issue. Giggle Academy, an educational service, was found to have reproduced hundreds of poems without permission, using AI to add illustrations and comprehension questions. Some instances involved sensitive content, such as poems about lynchings in the U.S., being paired with inappropriate, cheerful AI-generated imagery for young children. While some publishers successfully demanded the removal of content, debates continue over the application of "fair use" doctrines in the age of generative AI.