When AI Enters the Syllabus: Journalism’s Crossroads of Threat and Opportunity

AI in Journalism Education

AI is transforming how journalistic content is produced, distributed, and consumed (Bien-Aimé et al., 2025; Túñez-López et al., 2019). Consequently, journalism programs are under growing pressure to prepare students with skills and judgment necessary to engage with these technologies. However, prior research on AI-related curricular integration remains fragmented and uneven (Arzuaga, 2022). Tejedor et al. (2024) found that Spanish journalism programs offer little to no structured instruction on AI. This finding was echoed by Imran (2025) in analyses of curricula in Egypt and Australia, suggesting a significant gap between industry innovation and classroom preparation.

Seo et al. (2025), in a comparative study of journalism students in Brazil and the United States, found that while students broadly recognize AI’s professional relevance, their preparedness to adopt these tools varies widely. These differences may stem from curricula disparities, indicating that exposure, or lack thereof, to AI-related content may shape students’ preparedness. Similarly, Talaue (2023) highlights a mismatch between students’ initial intentions and their actual use of AI in their academic writing, underscoring the need for clear guidance. These gaps suggest that individual instructors often bear the responsibility of defining AI’s place in the classroom, leading to inconsistent implementation and student confusion.

Hossain and Wenger (2024) and Pavlik (2023) reinforce this concern, pointing to persistent gaps between institutional readiness and the pedagogical demands of AI integration. Hossain and Wenger (2024) found that while journalism educators widely recognize AI as a key force shaping the future of journalism education, many programs still lack coordinated policies or instructor training. However, Pavlik (2023) argues that journalism education must move beyond tool adoption to foster critical engagement with AI’s ethical and epistemological consequences. Pavlik (2023, p. 92) notes: “Educators should be considering . . . how to develop courses that train human students in the effective use of generative AI, as well as the threats it poses, including matters of ethics and potential bias.”

In a way, these studies suggest that the absence of structured AI instruction can lead to inconsistent usage, confusion over ethical boundaries, and uneven readiness. They also raise an important question; how are educators addressing these tensions through course design and policy language? This study, therefore, seeks to explore how syllabi present and regulate the use of AI in undergraduate journalism classroom. This leads to the study’s first research question:

Ethical Considerations of AI in Journalism

The use of AI in journalism raises ethical concerns that are not limited to academic integrity issues such as plagiarism or unauthorized use, but also extend to questions of accuracy, transparency and accountability in professional practice. Journalists, and by extension, journalism students, need to address concerns about algorithmic bias, misinformation, opacity, and a lack of accountability when automated systems generate errors (van Dis et al., 2023; Williamson & Prybutok, 2024; Zhang et al., 2020).

Whereas Perkins and Roe (2024) emphasize fostering students’ responsibility and reflection, Chaka (2023) approaches the issue from a technological angle, questioning the reliability of AI detection tools that often miss genuine misuse or wrongly flag legitimate work. For Chaka (2023), such tools risk creating an illusion of oversight without cultivating real ethical engagement. Johnson (2023) makes a related observation in reviewing syllabi, noting that many adopt rigid, prohibitive language–particularly in ethics courses–yet fail to invite students to critically consider the professional and moral stakes of adopting AI in the first place. Taken together, these studies point to the limitations of relying on punitive safeguards or rigid prohibitions: while they may signal oversight, they do little to cultivate the kind of ethical reflection and professional judgment that journalism education must foster in the age of AI.

This distinction between academic and professional ethics is crucial. While institutional policies may focus on preventing cheating or plagiarism, professional journalism demands a broader inquiry into authorship, transparency, and public accountability. Auman et al. (2020) found that although journalism instructors broadly agree on the core values students should learn, they diverge in how these values are taught. Some emphasized case studies, simulations, and role-playing to help students navigate dilemmas, while others relied more on lectures, abstract principles, or codes of ethics. Together, these variations underscore both the diversity of ethics instruction and the shared challenge of preparing students for the difficulties of a global, digital media environment.

Despite this growing discourse on AI ethics in journalism education, little research has focused on how course materials–particularly syllabi–address these concerns in practice. Do course materials offer students meaningful ethical frameworks, or do they treat AI as a tool to be banned or monitored without further reflection? This study addresses that gap by asking its second research question:

Constructionist Framing Perspective

Goffman (1974) argues that frames are schemata humans use to organize their experience in the social construction of reality. In journalism studies, framing is understood as a process of making some elements more salient while obscuring others in a communicating text to “promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation” (Entman, 1993, p. 52). Gamson (1989, p.157), however, contends that “a frame is a central organizing idea for making sense of relevant events and suggesting what is at issue.” Although somewhat different in their theoretical perspectives of framing, most scholars share similarities in their conceptualizations of framing function as making sense of social realities (Van Gorp, 2007).

Building on D’Angelo’s (2002) argument that framing research operates as a multiparadigmatic program encompassing cognitive, critical, and constructionist outlooks, this study situates itself within the constructionist paradigm. As D’Angelo (2002) explains, the constructionist paradigm is characterized by a “paradigmatic image of co-optation” in which journalists and communicators “create interpretive packages” that both reflect and shape the issue culture of a topic (p. 877). Within this view, frames are not static message attributes but social and discursive tools in which actors construct shared meanings. Accordingly, this study borrows the concepts of diagnostic and prognostic framing (Benford & Snow, 2000; Snow & Benford, 1988) but applies them interpretively to understand how journalism educators socially construct the meaning of AI, its impacts on journalism, hence justifying their policies, values, constraints, and expectations surrounding its classroom use.

Focusing on collective action, Snow and Benford (1988) offer several broad categories of frames including diagnostic, prognostic, and motivational. In this study, diagnostic framing refers to how instructors define or problematize AI integration in journalism education and attribute responsibility or constraints (Benford & Snow, 2000; Snow & Benford, 1988). Prognostic framing captures how they construct possible pedagogical, ethical, or professional responses to these perceived challenges. These framing dimensions are treated not as fixed message attributes but as interpretive acts through which educators articulate shared understandings of AI’s role in journalism education.

This orientation recognizes that instructors’ framings of emerging technologies are shaped by institutional norms, instructor agency, and pedagogical priorities. As Coatney (2022) shows in a study of journalism textbooks, materials often reflect conflicting visions of the journalist’s role, with some emphasizing traditional watchdog ideals while others promote technological adaptation or civic collaboration. This study extends that insight by turning to journalism course syllabi, which similarly function as institutional texts that not only guide instruction but also articulate normative expectations about the profession. The study explores how AI is framed in journalism syllabi, shedding light on how programs position emerging technologies in relation to journalistic ethics and professional identity.

Framing AI Through Syllabi

While ethics are central, the framing of AI in course documents may ultimately shape how students engage with the technology. Syllabi play a complex role in journalism education. They operate simultaneously as administrative tools and pedagogical guides that communicate instructor values, institutional priorities, and assumptions about professional practice (Doolittle & Lusk, 2007; Fuentes et al., 2021; Johnson, 2023). While often viewed as “contracts” between instructor and student (Parkes & Harris, 2002), syllabi are also shaped by institutional cultures, accreditation standards, and risk-averse policies (Eberly et al., 2001). Such factors may discourage experimentation or critical engagement with emerging technologies.

Petrotta et al. (2024) show that syllabi often reflect dominant professional values, especially in rapidly evolving fields like sports journalism, where responsiveness and trend awareness are emphasized. Mesmer and Miller (2024) similarly found that instructor experience plays a significant role in determining whether syllabi engage topics such as press hostility or industry threats. These studies together demonstrate how course documents are used to frame ethics, professional identity, industry pressures, and perceived threats to journalism practice.

Instructors may frame AI as an opportunity for innovation, or a threat to journalistic integrity, or both. These framings matter, not only for how students understand AI in journalistic practice, but for how they develop their own ethical reasoning and professional identity. The classroom can be seen as a key site for constructing professional identity and norms (Richardson, 1994). At the same time, Aldridge and Evetts (2003) caution that classroom learning has limits when it is detached from the realities of professional practice. Johnson (2023) found that ethics syllabi often emphasize surface-level cognition, such as remembering or applying rules, instead of fostering critical and creative engagement with ethical questions. While many syllabi identify critical thinking and argumentation as central, Johnson (2023, p. 19) observed that they rarely connect these skills to students’ personal decision-making, creating what they call a “contradiction” between professional ethics and everyday ethical practice. This narrow orientation, the scholar argues, risks reducing ethics to professional compliance rather than cultivating reflection that extends beyond the newsroom. This study builds on that insight by asking how syllabi frame the professional and ethical role of AI, through both the problems they identify and the solutions they propose. This leads to the study’s final question:

Data Collection

This study focused on journalism programs accredited by the Accrediting Council on Education in Journalism and Mass Communications (ACEJMC). A total of 32 ACEJMC-accredited institutions were selected to ensure representation across geographic regions (Northeast, Midwest, South, and Western), and institutional types (large public universities and smaller regional institutions). Each program’s chair, director and administrators were contacted via email in spring and summer 2025 with a request to share syllabi from four core undergraduate course categories:

These categories were selected based on both curricula structure and conceptual relevance. Each course type represents distinct domains of journalism education, with differing instructional goals and likely relationships to AI. For example, foundational writing courses may emphasize originality and authorship, while data journalism courses may be more open to automation and computational tools. Ethics courses, by contrast, may offer spaces for critical reflection on the implications of GenAI in professional practice. This approach aligns with prior studies that underscore the pedagogical and epistemological distinctions across domains (Ames et al., 2019).

Of the 32 schools contacted, 15 participated in the study. Twelve institutions submitted syllabi directly via email, and additional syllabi were collected from the publicly accessible websites of three other institutions. While schools were asked to provide at least four syllabi, some institutions submitted only one, while others provided substantially more than four. From these varied submissions, syllabi were purposively selected for inclusion, resulting in a final data set of 60 drawn from 15 schools. The sampled syllabi came from courses held in the Spring 2024, Fall 2025, and Spring 2025 academic terms.

A complete list of participating institutions and course categories is provided in Table 1.

OPEN IN VIEWER

Table 1.

Syllabi Included in the Sample by Institution and Course Category.

Region	Institution	Foundational writing & reporting (F)	News & feature production (N)	Visual & data journalism (V)	Media ethics & law (E)
Northeast	Pennsylvania State University	1	1	1	1
Northeast	University of Connecticut	0	0	0	1
Northeast	Temple University	1	1	1	1
Midwest	Michigan State University	1	1	1	1
Midwest	University of Kansas	1	2	2	1
Midwest	Ball State University	1	1	1	1
Midwest	University of Nebraska-Lincoln	1	1	2	1
South	University of Florida	1	2	2	1
South	University of Georgia	1	1	1	1
South	University of North Alabama	1	1	1	1
South	Texas State University	2	1	1	1
South	Louisiana State University	1	0	0	0
West	University of Colorado Boulder	1	2	2	1
West	California Polytechnic State University	1	1	2	1
West	San Jose State University	0	1	0	0
Total		14	16	17	13

Note. The sample includes 60 syllabi from courses in Spring 2024, Fall 2025, and Spring 2025.

Data Analysis

The syllabi were analyzed thematically by one of the authors in multiple rounds following Braun and Clarke’s (2006) approach to thematic analysis. The process began with familiarization and open coding of AI-related excerpts. These excerpts were drawn from sections such as course descriptions, learning objectives, assignments, academic integrity statements, and AI or technology policies. Initial codes captured patterns in instructional emphasis, ethical framing, and rhetorical positioning. Codes were then reviewed, grouped, and refined into themes and subthemes.

Drawing on a constructionist framing framework (D’Angelo, 2002), the analysis incorporated Snow and Benford’s (1998) aforementioned concepts of diagnostic and prognostic frames. Diagnostic framing was coded when syllabi defined AI as a problem or risk (e.g., warning that “ChatGPT . . . routinely make[s] up facts and fake citations”). Prognostic framing was coded when syllabi proposed responses, such as prohibiting AI use outright (“using any sort of AI tools will result in a grade of zero”) or allowing it conditionally with instructor approval (“students may only use generative AI. . . if you have requested to use it and had that request approved”).

Beyond identifying diagnostic and prognostic frames, close attention was also paid to the language, particularly how AI was positioned in relation to core journalistic values, such as originality, transparency, and professional responsibility, and whether it was treated as a tool to be used, a topic to be discussed, or a risk to be managed.

Themes were developed inductively and iteratively, allowing patterns to emerge across course categories while remaining sensitive to contextual variation, such as different course types. This approach enabled a grounded interpretation of how AI is framed in journalism syllabi, consistent with the constructionist lens guiding the study. All syllabus excerpts are anonymized and labeled according to course type. Each syllabus was assigned an internal identifier consisting of a category code and sequential number (e.g., F1—F10 for Foundational writing and reporting courses; N1—N8 for News and feature production; V1—V9 for Visual and data journalism; E1—E10 for Media Ethics and law courses). Multiple excerpts drawn from the same syllabus share the same label.

Theme 1: AI as a Threat to Learning and Professional Standards (Diagnostic Framing)

Across many syllabi, AI was framed as a direct threat to the core purposes of journalism education. These diagnostic frames positioned AI as plagiarism, academic dishonesty, or a shortcut that undermines learning and professional formation. Instructors emphasized originality, integrity, and accountability as non-negotiable values that AI use was seen to erode, particularly in foundational writing and reporting, and multimedia production courses. This theme addresses RQ1, RQ2, and RQ3, showing how AI is defined as both an instructional and ethical problem.

Theme 2: AI as a Tool Permitted Under Strict Boundaries (Prognostic Framing)

While many syllabi took a zero-tolerance stance on AI use, others acknowledged its growing role in students’ workflows (RQ1, RQ2). In these cases, instructors drew clear boundaries between acceptable support and unacceptable substitution. AI was permitted only for limited, mechanical tasks–such as grammar checking, transcription, or idea generation–while being prohibited from replacing core journalistic work. Instructors framed such use as something to be managed, not freely adopted, reflecting an effort to recognize the reality of GenAI without surrendering pedagogical authority or ethical standards. This framing, however, indicates the uncertainty across journalism classes about how AI should be incorporated in class work.

Theme 3: AI as a Subject of Ethical and Professional Inquiry

A smaller but notable set of syllabi positioned AI not only as a tool to regulate but as a subject of reflection and debate. In these cases, instructors invited students to consider AI’s ethical, professional, and social consequences, treating the classroom as a site of inquiry rather than compliance. This framing moved beyond prohibitions or narrow allowances, engaging students in deeper questions about journalism’s role in an AI-driven media environment.

Theoretical and Practical Implications

The inconsistency in framing across classes has instructional implications in journalism education. While instructors often make intentional choices, the absence of shared guidance across courses may lead to uneven student experiences across a journalism curriculum. Syllabi that emphasize strict prohibition often focus on preserving originality, authorship, and learning integrity. They express concern that AI bypasses the cognitive effort students need to develop as writers and reporters. This mirrors broader worries raised by van Dis et al. (2023), Williamson and Prybutok (2024), and Zhang et al. (2020), who argue that over-reliance on automation may weaken students’ critical thinking. Similar tensions have been observed globally. Tejedor et al. (2024) and Imran (2025), for example, show that journalism curricula in Spain, Egypt, and Australia often lack structured AI training, producing similar anxieties about erosion of professional standards.

At the same time, the findings show that some instructors allow limited or guided use of AI, specifically for brainstorming, grammar checks, or research support. This shows a more balanced stance, where AI is acknowledged as part of students’ workflow but tightly controlled. These instructors align more closely with Pavlik (2023), who advocates for responsible, transparent use of AI, in which one trains journalists instead of issuing a blanket rejection. Framing AI in this way reduces confusion and also prepares students for a newsroom environment where AI is already integrated into reporting workflows (Juneja & Mitra, 2022; Munoriyarwa et al., 2023). Wenger et al. (2024) and Pavlik (2023) emphasize that technological fluency is no longer optional; it is part of professional readiness. Seo et al. (2025) support this claim, showing that students’ preparedness for AI varies widely depending on curricular exposure, highlighting why structured, guided use is necessary for building competence.

However, when AI is treated solely as a disciplinary issue, something to be banned, punished, or hidden, students are given few opportunities to engage with its ethical complexity. This echoes Johnson’s (2023) critique that syllabi often impose rigid rules without offering students space to interrogate their meaning. The result is a surface-level ethical engagement, where students comply with rules but fail to develop ethical reasoning. Some syllabi in this study stood out by embedding AI into classroom discussions, readings, and case studies. These examples reflect what Perkins and Roe (2024) call for, moving beyond technical safeguards toward fostering ethical inquiry and responsibility. Talaue (2023) similarly shows that without such guidance, students often improvise their use of AI in ways misaligned with course expectations.

The role of the instructor also emerged as key. Without consistent institutional frameworks, decisions about AI use are left to individual instructors, leading to varied enforcement and messaging. Some act as strict gatekeepers, while others take a more facilitative role. Syllabi that emphasize instructor approval and disclosure suggest a model of negotiated ethical use, where students must justify their reliance on AI and reflect on its appropriateness. This approach models ethical judgment rather than enforcing blanket compliance. Yet instructor discretion can also produce disparities in student experience across programs. This aligns with Seo et al. (2025), who found that student preparedness for AI is shaped by exposure, or lack thereof, within coursework. It also reflects what Hossain and Wenger (2024) describe as a gap between institutional recognition of AI’s importance and the absence of coordinated training or policies for faculty. Such efforts could offer guiding principles that allow instructors to innovate while minimizing contradictory expectations across courses.

These disparities underscore that syllabi reflect not only course aims but also instructors’ professional values and identities. Mesmer and Miller (2024) and Petrotta et al. (2024) emphasize that instructors’ professional experiences strongly shape how challenges like AI are addressed in classroom texts. This study similarly found that how AI is framed in syllabi reveals deeper assumptions about journalism’s future and educators’ role in preparing students for it.

More broadly, diagnostic and prognostic frames were evident throughout the syllabi. Diagnostic frames portrayed AI as unreliable, inaccurate or unethical, particularly when linked to generative tools that “make things up” or obscure their sources. Prognostic framings varied, ranging from strict bans to carefully bounded allowances to curricular integration. These framings reflect different visions of journalism education: one that protects traditional skills, one that cautiously adapts to change, and one that encourages innovation within ethical guardrails. As Hossain and Wenger (2024) and Pavlik (2023) argue, these tensions highlight the need for deliberate, coordinated approaches to AI in journalism education rather than fragmented, instructor-by-instructor decision-making.

Limitations and Future Research

This study is limited by its reliance on syllabi, which may not capture how policies are implemented in classroom practice. Policy implementation could differ from syllabus language, particularly because of the speed with which AI-related technology evolves. Future studies could examine assignments, lecture slides, and classroom activities to see how AI discussions unfold in real time. Similarly, while the analysis offers insight into instructors’ intentions, it cannot speak directly to outcomes; we do not know how students interpret or respond to these framings. Future studies could address this by interviewing professors about their pedagogical strategies and surveying students about how they experience and apply AI-related guidance. Future work could also more directly look at elements of the syllabus that recommend, require, or ban specific AI tools for use in journalism activity.

The sample also reflects programs willing to share syllabi for this study, which might have highlighted institutions actively engaging with AI. However, non-responding schools may frame AI differently. Future studies should expand data sets to test representativeness. More so, focusing on U.S. ACEJMC programs provides comparability within a shared accreditation framework, but it limits regional and global insight. Future cross-national studies could reveal whether similar or alternative framings of AI are emerging elsewhere. Also, since the sample consists only of public institutions, the findings may not reflect how private journalism programs frame AI in course syllabi. Institutional governance, resources, and policy environments may differ in private settings, potentially shaping alternative approaches to AI integration. Future studies could address this by sampling this population. We also did not specifically analyze syllabi by university; future studies could look at institutional differences between syllabi to more specifically assess whether universities are internally consistent with their AI policies.

This study focused on courses in journalism practice, rather than courses in communication theory and research methods. Because this study examines how AI is framed in relation to authorship, integrity, verification, and newsroom-related skills, we focused on course categories where these issues are most salient. Future research could consider more research-focused courses, as they might conceptualize AI in distinct ways.