Higher Ed faculty spend a significant portion of their workload on assessment. Time on assessment often ranges from 11% to over 35% of their weekly hours, mostly on grading and providing feedback for paper-based assignments. Studies suggest this averages roughly 5 to 15 hours per week for undergraduates (Hardison, 2022). The time commitment varies heavily based on undergraduate or graduate coursework, assignment type; grading papers can take 20–60 minutes per student, while 80, 3-page papers might take over 80 hours total with intensive writing courses or large classes leading to much higher, sometimes unsustainable, grading loads (Mason, 2023). 

     The critical thinking churn of both assessment and written artifacts can become repetitively mundane for both faculty and their students. As undergraduate and graduate students confront the reading and writing demands of course work, there is a temptation to use shortcuts which tamper with authentic writing, critical thinking, concept acquisition, and personal transparency with their teachers. Sometimes the problem is not the AI use ~ it is the avoidance of personal reflection and the prioritization of time for critical thinking. Many papers, especially in graduate courses, require students to connect philosophy, policy, or research to their own teaching and learning.  AI is unable to do this work with authenticity. Students who outsource the critical thinking churn and accountability work skip the actual intellectual mental and reflective exercise the assignments are designed to create. The shortcut practice of using AI shortcuts occurs even when faculty build a fair and transparent AI policy.

     A new complicating factor for students who use of artificial affordances in place of their authentic voice is a new characteristic of AI usage known as ‘Double Flag’ Attribution Flip is alerting on university detechtion software checking for plagiarism and AI usage.

AI Match and ‘Double Flag’ Attribution Flip:
      In an interesting new development, when students reuse AI generated text in other assignments, there is a resulting higher attribution percentage. When the ‘source’ text being quoted was also AI-generated, the Copyleaks report will flag it as a misattribution or ‘source match.’ Essentially, you cannot ‘own’ an AI output by quoting it. Quoting a machine, for example AI, and labeling the resulting text as your own present or previous writing is misleading (Campbellsville University Libguides, 2026; Library and Learning Center, 2026) and objectively goes against most course requirements for an authentic human voice (Sharpe, 2025; Tufts University, 2026).

     However, if that earlier work was AI-generated, the AI-generated content creates a technical “double flag” in the attribution software such as TurnItIn and Copyleaks (De Amicis, 2026). The system recognizes the underlying machine patterns in the text (Campbellsville University Libguides, 2026) and can flag as both AI and traditional plagiarism. 

As a connecting reference, here are general AI tips and ideas from my recent AI presentation for the LOPES conference.

                                                                  References

Note: Post originally published April 21, 2025. Updated April 16, 2026 with updated resources and references.