"Young Scholar[s] on the Beat": Multimodal Composition as a Form of Critical AI Literacy Pedagogy

Synthesis: Burriss, Eeds, Smith, Ziegler, Villanueva, and Deweese (2026) argue that AI literacy should be reoriented around creative, collaborative composition rather than narrow technical competencies, and present a classroom study in which 22 eleventh-grade students composed video public service announcements about AI Ethics topics of their choosing. Grounded in multimodal composition and Critical Posthumanist Literacy, the unit asked youth to translate abstract principles — surveillance, data consent, algorithmic accusation — into emotionally resonant films using sound, image, text, and their own bodies. Analysis of seven PSA videos, 18 end-of-unit surveys, 12 daily reflections, and classwork found that students overwhelmingly portrayed AI negatively, especially school-based surveillance and punitive AI-driven accusation; depicted harm as emerging from tangled human–machine responsibility rather than from a villainous tool; and still closed with resilient messages of hope and Learner Agency. Students described composing as joyful, creative collaboration and reported that the project deepened their understanding of AI ethics, with 15 of 18 responses affirming new learning. The authors position storytelling and advocacy — resistant to traditional assessment — as core to critical AI literacy pedagogy that centers student identity, choice, and voice.

Key Findings

Negative portrayals of AI dominated: Most groups chose risk-and-harm narratives featuring surveillance, punishment, and lack of informed consent around school-regulated laptops, electronic "hall passes," and plagiarism-detection programs. One group explained, "We picked AI privacy in schools because it was personal to all group members."
Human–machine entanglement, not simple villainy: Across all 7 films, harm emerged from human–AI interaction rather than the tool alone — "bad actors" included a principal misusing an AI tool, a deepfake banking scam, and an anthropomorphized "AI stalker." Anthropomorphized AI characters played by human actors appeared in 3 of 7 films.
Persistent hope and agency: Despite tense music and dire slogans, the films ended on agency — "the solution is in our reach" (Failsafe? I Feel Safe!) and a student freeing themself by signing a Bill of Rights (AI Stalker), with the AI Rap lyric "privacy and learning hand in hand/it's up to us to take a stand."
Students flipped the cheating narrative: In The AI Accusation Chronicles, the ethical issue centered was not student cheating but adults' overreliance on faulty AI determination; the plot is resolved when a human witness vouches for the accused student.
Composition felt joyful and collaborative: One student called the project "Unequivocally, one of my favorite activities that we have done at the [program] thus far," citing cross-group casting, acting, peer assessment, and the closing "red carpet premiere."
Students reported AI-ethics learning: Of 18 end-of-unit survey responses, 15 were positive about how making the video changed their understanding of AI ethics; the 3 negative responses cited prior knowledge of these issues.
Youth believed in their own impact, with dissent: 14 of 18 students said youth can influence issues like AI, but one student, Harris, countered that youth "lack the ability to vote" and "completely lack any kind of credibility."
Technical clarity was sometimes sacrificed: "Informed consent" was translated accessibly as "We gotta approve," yet some terms (e.g., "failsafe") were never explained — prompting the authors to plan deeper conceptual grounding in future iterations.

Study Design & Method

Context and participants: 22 eleventh-grade students (41% Black/African American, 23% White, 18% Asian, 14% Latinx, 4% Native American; 50% female) from an urban public school district, enrolled in a 4-year, application-based, credit-bearing co-curricular STEM research program that meets one day a week at a nearby university.
Unit and task: Researchers adapted the existing "Idea to Film" module; 7 student groups produced 90-second to 3-minute video PSAs on an AI ethics issue of their choice, supported by a "video planner" slide deck, a PSA genre study, and discussion of the White House Blueprint for an AI Bill of Rights.
Theoretical framing: Critical Posthumanist Literacy (ontology, agency, ethics & justice, Pedagogies and Teaching Strategies) combined with social semiotics and multimodality; the kineikonic (moving video) mode anchored the multimodal transcripts.
Data sources: 7 final PSA video files; 18 end-of-unit survey responses and 12 daily reflections; 4 chart papers and 4 video planners/scripts collected from teaching and classwork.
Analysis: Iterative open coding by three education researchers with whole-team thematic discussion, re-coding of videos and written responses against the research questions, plus multimodal transcription and content analysis of co-occurring modes (visuals, text, sound, speech).
Ethics and funding: IRB-approved study with student assent and parental consent, pseudonyms for all names, and support from National Science Foundation award DRL-2112635; raw data are not shareable for confidentiality reasons.
Limits: A low-stakes, exploratory, single-cohort design with no formal summative assessment, limited time and materials, and a setting that already had strong collaboration norms and student video expertise.

What this means for practice

Instructors. Assign low-stakes multimodal composition — a short video PSA on an AI ethics issue students choose and address to a real audience — because 15 of the 18 end-of-unit survey responses said making the video deepened their understanding of AI ethics; the work needs teacher openness rather than AI expertise, since joint investigation of a specific system surfaced and addressed Misconceptions about AI.
Instructors. Anchor the unit in students' own encounters with AI, such as school surveillance, electronic hall passes and plagiarism detection, which is where this class located the ethical stakes.
Instructors. Develop collaboration norms and inventory student expertise in videography and editing at the outset, since this class already had both and classrooms without them will need to scaffold them deliberately.
Instructors. Substitute lower-overhead products — infographics, short presentations, graphic-novel panels or mini-podcasts — when time or materials make live-action film impractical.
Researchers. Assess AI literacy with collaborative and multimodal evidence such as reflections, artifacts and civic discourse, since existing AI literacy scales assume individually measurable performance.

Limitations

The study rests on a single cohort of 22 eleventh-grade students in one urban district's application-based STEM research program, with no formal summative assessment of what they learned.
The classroom already had strong collaboration norms and substantial student video expertise, so the unit's feasibility in ordinary classrooms is untested.
The evidence is small and largely self-reported: 7 groups' final videos, 18 end-of-unit survey responses and 12 daily reflections.
The researchers' own instruction shaped what students made — the AI Bill of Rights discussion and the PSA genre study influenced topic choices — and some technical terms (e.g., "failsafe") were never explained, which the authors concede.

Connected Concepts

Connected Articles

A Posthumanist Approach to AI Literacy — Posthumanist reframing of AI literacy theory and practice
Reorienting Critical AI Literacy: A Community-Rooted Praxis of “Resisting AI” — Reorienting critical AI literacy toward critique and power
Students' Perceptions of Multiliteracies Development Using AI-Assisted Portfolio Assessment — Multiliteracies and AI portfolios in youth learning
When Youth Enter The Chat: An Epistemic Shift in the Validation of LLM-Based Measures of Student Talk — How youth talk with language models in everyday learning
Multimodal Learning with Generative AI — Generative AI in multimodal learning contexts

Citation

Burriss, S. K., Eeds, A., Smith, B. E., Ziegler, H. H., Villanueva, A., & Deweese, M. (2026). "Young Scholar[s] on the Beat": Multimodal composition as a form of critical AI literacy pedagogy. Journal of Adolescent & Adult Literacy, 70(2), e70064.

Metadata

Study design	Thematic Analysis
Discipline	English Education
Level	Secondary
Intended audience	Instructors, Teachers
Learning and instruction	Creativity, Critical Pedagogy, Storytelling in Education, Student Engagement
Technologies and techniques	Multimodal AI
Research methods and evaluation	Qualitative Research
Equity, ethics, and responsible use	Ethics
Foundations of AI in education	Learner Agency, AI Literacy

Source History Edit Chat