Beyond the Trap: Why Hidden AI Prompts Fail Educators and Students episode artwork

EPISODE · Jun 18, 2026 · 45 MIN

Beyond the Trap: Why Hidden AI Prompts Fail Educators and Students

from Deep Learning Dialogues

In this episode of Deep Learning Dialogues, hosts Whitney McKinlay and Katrina Gouett sit down with Dr. Adam Pacton to unpack the controversial practice of "prompt injection" or hiding invisible text in assignments to catch students using generative AI. Moving past the initial "ick" factor or ethical intuition that such tactics turn classrooms into policing environments rather than spaces of mutual trust, the conversation explores the concrete legal, technical, and pedagogical failures of these traps. Dr. Pacton explains how hidden instructions act as an accessibility nightmare under regulations like the ADA Title II and Ontario's AODA, because screen readers accidentally surface the invisible text, confusing students who rely on assistive technology. Rather than relying on flawed detection mechanisms, the episode advocates for a systemic paradigm shift toward "process over product," urging institutions to provide teachers with the cover and resources needed to meaningfully redesign assessments and foster relational intelligence. You can read the episode summary here.Dr. Adam Pacton is an associate teaching professor and Dean's fellow for AI literacy and integration at the College of Integrative Sciences and Arts at Arizona State University. Working at the intersection of pedagogy, ethics, and emerging technology, he focuses on institutional AI governance while helping educators move away from fear-based compliance toward thoughtful, human-centered classroom integration. With a PhD in rhetoric and composition, Dr. Pacton brings a deeply grounded perspective to language practices, change management, and student agency in the age of artificial intelligence. Outside of his academic and administrative leadership roles, he rounds out his human perspective as a father, an avid video game player, and a passionate fan of fantasy and sci-fi literature.Episode Link:-ASU Guidance on Hidden-Text/Prompt Injection Techniques in AssignmentsWant to know more?You can check out our: WCDSB GenAI Guidelines, infographics, and Innovation website: https://innovate.wcdsb.ca/Feedback? You can ask your questions or give us feedback on the show here Want to get in touch? Contact Katrina & Whitney by email at: [email protected] and [email protected] or on LinkedIn Hosted on Acast. See acast.com/privacy for more information.

Episode metadata supplied by the publisher feed · Published Jun 18, 2026

Embed this episode

Ready to play

Beyond the Trap: Why Hidden AI Prompts Fail Educators and Students

0:00 45:53

No transcript for this episode yet

We transcribe on demand. Request one and we'll notify you when it's ready — usually under 10 minutes.

No similar episodes found.

Frequently Asked Questions

How long is this episode of Deep Learning Dialogues?

This episode is 45 minutes long.

When was this Deep Learning Dialogues episode published?

This episode was published on June 18, 2026.

Can I download this Deep Learning Dialogues episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.
URL copied to clipboard!