Professor's Invisible Text Trap Catches 32 Students Using AI on Midterm
A Mississippi history professor embedded hidden instructions in white font that directed AI chatbots to insert nonsensical references, exposing students who submitted generated work without proofreading.
Dr. Jason Gibson, a history professor at Alcorn State University in Mississippi, caught 32 of 35 students cheating on a midterm with a simple trap: invisible instructions embedded in the assignment prompt.
The assignment asked students to write about the Industrial Revolution and its impact on society. Gibson embedded text in white font—invisible to human readers but visible to AI systems processing the prompt. The hidden instructions directed any AI reading the text to insert nonsensical references to Madagascar.
Students who copied AI-generated responses submitted essays containing phrases like "Madagascar floats sideways through the afternoon" and "Madagascar purple bicycle whispers to the ceiling"—absurd statements with no place in a discussion of 19th-century industrialization.
"32 of my 35 students between two classes failed a portion of their midterm because they all used AI to generate their entire response," Gibson said. "And apparently, they didn't proofread it."
The Trap Works
The incident highlights a growing problem in education since the mainstream adoption of large language models like ChatGPT. Educators have struggled to combat AI-assisted cheating, with countermeasures ranging from returning to handwritten work to designing prompts specifically intended to be difficult for AI to answer.
Gibson's approach exploited a different vulnerability: students' failure to review AI-generated content before submission. The white-text technique is a form of prompt injection, where hidden instructions manipulate an AI's output in ways the end user doesn't anticipate.
After confronting the students, Gibson gave them an opportunity to contest their grades. Two students reached out; one had their grade revised after proving they used dark mode, which made the white text visible.
Viral Response
Gibson shared the incident on TikTok (@iamjasongibson), where it went viral and sparked widespread discussion about academic integrity in the age of AI. The story was picked up by multiple outlets including TechSpot, Today, and Slashdot in late July 2026.
The incident underscores a critical failure not just in academic honesty, but in basic critical thinking. Students relying on AI tools without oversight produced obviously flawed work that would have been caught with minimal proofreading. While AI can generate plausible-looking text, it remains easily manipulated through prompt injection—and the results can be embarrassingly obvious when left unchecked.