Preparing AI-Ready Engineering Talent: Faculty Perspectives on Generative AI, Foundational Competence, and Workforce Readiness

Synthesis: Sadik (2026) reports a basic qualitative interview study of ten faculty members in an AI-focused engineering department at a U.S. public university, analyzed by inductive thematic analysis. Four themes describe generative AI as an inevitable but questioned learning tool, a source of tension between performance and demonstrated learning, a catalyst for rebalancing engineering curricula, and an uncertain influence on workforce readiness. Participants saw opportunity in coding, explanation, feedback, and ideation, yet disagreed about when GenAI should be introduced and whether traditional programming requirements should change, and all ten raised integrity concerns. Conditional integration is the study's central contribution: acceptable use turns on the learner's developmental readiness, the task's purpose, the ability to verify outputs, disclosure, and accountability for failure.

Key Findings

1. Every participant expected AI to reshape engineering work, but none agreed how. Forecasts ran from change that would be "enormous" (P10) to engineering that would "survive to some degree" (P6); three named coding and writing as near-term vulnerable.
2. Ten faculty in one department reached saturation, not consensus. Of 27 eligible faculty invited, ten completed interviews of 29 to 52 minutes; by interviews eight and nine new data reinforced existing themes and the tenth added none.
3. Generative AI was inevitable but not endorsed. Nine participants reported students already using GenAI for coding and writing; three called it a black box and four said students rarely cited AI-generated material.
4. Better artifacts did not prove better learning. All ten raised academic-integrity concerns and eight worried about effects on learning, critical thinking, Creativity, or foundational skills; P6 found reports "much better" without evidence of better skill sets.
5. Faculty split on whether programming remains foundational. P6 asked "What's the point of learning how to program for that specific hardware?" while P9 held that good program designers must first understand fundamentals learned through "that drudgery work."
6. Nine participants wanted curricular revision, not a GenAI course. Seven emphasized AI literacy, six ethics, six revisiting learning objectives, and four infusing AI learning across existing disciplinary courses.
7. Five conditions governed acceptable use: developmental readiness, the task's purpose, the ability to verify the output, disclosure and ethical provenance, and professional accountability for failure.

Conditional integration, not permission or prohibition

Participants did not divide neatly into supporters and opponents; their judgments turned on whether GenAI extended a learner's capability or replaced the reasoning a task was designed to develop. P1 accepted the tool for seniors who had already learned several programming languages but warned that first-year students would "generate the script" before understanding why a loop was needed. P7 permitted GenAI for parts of technical work yet discouraged it for discussion pieces meant to show students' own interpretations, and P9 welcomed the shift from code production toward code design while keeping "ultimate authority" over what should be implemented.

Productivity versus demonstrated learning

Offloading the work risked hiding whether students could explain, verify, or reproduce the underlying reasoning. P6 cautioned that "the quality of the reports is much better" but this "doesn't necessarily mean that the students developed better skill sets," and P5 described a student who "didn't run the code to see that it is working." Instead of detecting tool use, participants proposed Assessment that demands reasoning, process demonstration, output verification, and oral defense, with P10 requiring students to stay "responsible for every line of code."

Rebalancing engineering curricula for AI-ready talent

Nine participants called for curricular revision, and their recommendations did not amount to replacing engineering foundations with a GenAI curriculum. Seven emphasized AI literacy — how systems are built, which tasks they can perform, where they fail, and how outputs should be evaluated. Six emphasized Ethics, extended from plagiarism to disclosure, provenance, data Privacy, bias, safety, and downstream responsibility; P10 insisted, "You're not relieved of responsibility" because a result came from machine learning. Five emphasized higher-order problem solving and three communication: P7 proposed work around "higher-order problems that might integrate more across domains," while P10 argued that engineering work would require more dialogue with these systems. Six also wanted learning objectives revisited — a program-level coordination problem, not a course fix.

Workforce readiness under conditions of uncertainty

All participants expected AI to influence engineering work, and the study treats their differing forecasts as predictions rather than evidence about employment. Three named coding, writing, and information work as especially vulnerable in the near term; P10 anticipated fewer humans in some decision loops, while P8 expected "low-level coding" to be automated but held that work "beyond the expectation" still needs people: "Only humans can do that." Seven identified AI literacy as important for future engineers and four stressed the continuing importance of foundational engineering knowledge; P2 compared AI knowledge to mathematics, calling it "like a math skill for engineers." Their profile for professional preparation combines tool use with evaluation, communication, complex problem solving, lifelong learning, and responsibility.

What this means for practice

• Instructors. Stage GenAI by developmental readiness and verification capacity rather than banning or permitting it outright: P1 supported it for seniors who already knew several programming languages but warned that first-year students would generate scripts before understanding why a loop was needed.
• Instructors. Require evidence of reasoning — explanations of process, output verification, oral defense — and keep students "responsible for every line of code" (P10) rather than trying to detect tool use; disclosure makes the work theirs to defend.
• Curriculum designers. Coordinate revision at the program level so AI literacy, ethics, communication, and higher-order problem solving recur across courses while foundational practice stays visible: nine of ten participants wanted revision, seven named AI literacy, six ethics.
• Faculty developers. Build capacity for defensible pedagogical decisions under changing tools: five participants described uncertainty about acceptable uses, policies, and assessment design, and P4 reported no developed rubric for careful tool use.
• Administrators. Write policy around durable principles — disclosure, verification, task purpose, accountability — rather than around a specific tool, since participants expected tool-specific boundaries to age quickly.

Limitations

• Ten of the 27 eligible faculty in one AI-focused engineering department at one public university completed interviews, so departmental culture and disciplinary concentration may have shaped the reported views.
• Findings rest on faculty perceptions, not observations of classroom implementation or direct measures of student learning and workforce outcomes.
• The manuscript names no specific model or version and gives no data-collection window, so participants' judgments cannot be tied to a particular generation of systems.
• Credibility checks were modest — one peer-debriefing session and member checking with three of ten participants — and the full dataset cannot be shared under the study's IRB approval.

Citation

Sadik, O. (2026). Preparing AI-Ready Engineering Talent: Faculty Perspectives on Generative AI, Foundational Competence, and Workforce Readiness. International Journal of AI in Pedagogy, Innovation, and Learning Futures.