Join/Renew
Recruiters are being flooded with AI-generated materials that make it difficult to distinguish real from fake candidates. We need to teach our students to use AI properly and ethically, in a way that doesn’t deprive them of their own voices and won’t make them blindly reliant on copy and pasting AI outputs.
A Crisis Hidden in Plain Sight
I witnessed three instances within a week. Three different students. Three different majors, career objectives, and individual histories. Yet the three generic resumes were so alike that I had to double-check that I was looking at different files.
The students had not collaborated.
There had been no conscious copying of one another on their part.
They had, however, all used a simple AI prompt to create a resume and then copied it to a Word file and submitted it as their own.
The initial red flag was triggered for me during my first advising appointment of the week. I asked the student to share what she was most proud of in her volunteer experience at a homeless shelter, something that I spotted on her resume when scanning it earlier. She hesitated and seemed uncertain about how to respond. Then she said something that startled me; she asked me to tell her what her resume said, as she had not written that part herself.
On further questioning, she explained that she had created her resume using AI. I had to pause and take the time to explain to her the dangers of directly copying and pasting from AI, how AI can itself can imagine having a credential and hallucinate information, how it could never experience the same thing as her, and that the voice in the resume was not her voice or an accurate reflection of her skills or experience. Because of this, she would never be able to defend it, at least not in a face-to-face interview.
I was not concerned that she had used AI. I was concerned because she didn’t understand the distinction between using AI as a co-pilot and putting AI in the driver’s seat and trusting it to navigate for her.
I was no longer surprised at the third appointment of that week. I was, however, looking at a deeply concerning pattern, and I was alarmed.
It was not just the career coordinator part of me that was concerned. I was also worried as a human resource professional with more than 20 years of experience. The job economy that these students were entering is not forgiving of such a lack of connection or authenticity. Recruiters are not seeking ways to make saying yes to a candidate easier. They are more often looking for quicker ways to weed out candidates who aren’t qualified or verified. Moreover, the 2026 talent market has provided recruiters with a flood of AI-generated materials that make it increasingly harder to discern whether candidates are real or not.
The scale of the AI problem is well documented. A 2025 Greenhouse report surveying more than 4,100 job applicants, recruiters, and hiring managers across four countries found that nearly half of U.S. job seekers (49%) submitted more applications in the past year than the year before, with Greenhouse CEO Daniel Chait describing talent acquisition teams as drowning in application volume.1 The same report found that 34% of recruiters reported spending up to half their workweek filtering spam and junk applications.2
AI auto-apply platforms have intensified this problem. There are tools that post hundreds of applications on a candidate’s behalf in a single sitting, with AI dynamically tailoring resumes and cover letters for each posting. The stated objective is efficiency. The practical result is a job market so flooded with homogenous and ghostwritten applications that a genuinely qualified candidate can barely surface.
Then there is the fraud dimension. Ninety-one percent of U.S. recruiters report having spotted candidate deception of some form, and 65% of hiring managers have identified applicants appearing to cheat with AI, reading from scripts, hiding prompt injections in resumes to bias applicant tracking system scoring, and, in a smaller number of cases, appearing as a deepfake during a video interview.3
This concern is not confined to one survey. A separate study of 3,000 American managers found that 59% have personally suspected a candidate of using AI to misrepresent themselves during hiring, and nearly two-thirds believe that job seekers are currently more skilled at faking it with AI than recruiters are at detecting it.4 Gartner has estimated that by 2028, one in every four candidate profiles globally will be fabricated.5 In late 2025, Daniel Chait stated that trust is at an all-time low for both job seekers and recruiters.6 Career services professionals need to sit with that sentence.
It is easy to read this landscape and conclude that AI is the problem. That is the wrong conclusion. AI is already embedded in the employer side of hiring. The students we serve are entering a world of work that runs on these tools. Teaching them to avoid AI is not only impractical, but it is also a professional disservice. It is not a question of AI use. The issue lies in AI use in the absence of responsibility, lack of critical assessment, and the authentic voice of the student that grounds all outputs.
A student has not hacked their job search when they submit a resume that has been constructed by AI without any human involvement. Instead, they have created a report that they cannot defend, one that could be filled with delusional or generalized qualifications (making them sound like all the other applicants), and which falls apart as soon as the recruiter or hiring manager poses a question or two. This homogenization of ghostwritten candidates creates professional camouflage. Applicants look and sound like everyone else, and in a hiring environment that is already programmed to resent AI-generated content, it is a liability.
The Authenticity Gap
I have heard this issue called the authenticity gap, and the name perfectly describes this phenomenon. The difference between what is submitted by the student and what is justifiable in a real-life professional dialogue is the gap. This is not a new circumstance to career services—advisers have always dealt with students who exaggerate qualifications or borrow language that does not quite sound like them. What AI has done is widen the gap to a level that is very different from anything that the field has seen previously.
Typically, a student who is writing their resume on their own has read it more than once, has critically thought about what to include or leave out, and can expand with details and vivid stories about how they did what they did in every experience listed on their resume. In comparison, a student who pasted an AI output into a Word document without reading the output critically has produced a document that is property-free. There is no story. There is no authentic person behind it who could stand in a room and share in detail: This is who I am.
Employers notice this gap. Once they pose a follow-up question, they can detect the disappearance of the authentic human. They can notice the tone shift from resume to interview performance. They can identify LinkedIn profiles that aren’t relevant to the individual on the other side of the interview table. They can tell in cover letters that sound cold, formulaic, and generic. The tell is this: The lack of authentic human connection and voice highlights that no human is behind the document. Unethical AI use is not a formatting issue. It is an issue of professional credibility. Career services practitioners are perfectly positioned to help address this issue with ethical AI use and the preservation of authentic voice.
From Prohibition to Pedagogy: Why Policing AI Is the Wrong Answer
Prohibiting the use of AI has been the main institutional reflex of the higher education sector. We use AI-detection software, academic integrity policies are restructured to ensure that AI use is identified, and assignments are designed in such a way that AI assistance is impossible.
It is also the wrong response, and in career services, it is actively misaligned with the world students are entering. Seventy-eight percent of businesses now apply AI to at least one business function, up from 72% in early 2024 and 55% the year before.7 Some applicant tracking systems rank candidates before a human reads a single word.8 Building walls around AI use does not give students AI fluency. It teaches them to see AI as a threat to evade rather than a tool to use with intention.
The question was never whether students would use these tools. That question was settled the moment AI assistants became free and embedded in everyday platforms. The only question that matters is whether students use them with accountability and critical judgment. According to a recent 2026 NACE survey, 55% of career centers are offering workshops on AI-assisted job searching and another 20% plan to do so.9 The professionals best positioned to respond are not those building higher walls. They are those building better frameworks.
Developing the Framework: GENS 401
The question that our team at the University of Southern Indiana (USI) Career Success Center had to address was not to teach students not to use AI, but how to teach students to use it in such a way that would not deprive them of their own voices, and would not make them blindly reliant on copy and pasting AI outputs.
Our solution was the development of GENS 401: Career Readiness and Professional Development, a one-credit, eight-week, asynchronous course that is offered fully online and serves juniors and seniors.
Introduced in fall 2025, GENS 401 is designed to teach students about how to use AI and assess AI products critically and about their personal responsibility for the implementation of AI-assisted work. The fall 2025 class consisted of 15 students; there were nine students in the spring 2026 class. The students’ majors, ages, and career goals varied. Between terms and based on results with the fall 2025 class, the course was adjusted: The number of words for self-reflections was modified and the expectations connected with the specifics of the AI integration were also recalibrated to reflect the actual measures which were taken by students in real time. Each course offering of GENS 401 is shaped and transformed by the experience and output quality of the previous cohort, making this a living curriculum rather than just a fixed product.
The GPS Principle
Each meeting in GENS 401 is based on the GPS principle: AI is the co-pilot. The student is the driver.
This framing does specific and significant work. It does not position AI as a rival. It doesn’t frame AI as a shortcut. It puts AI in the hands of a student who is aware of where he or she is headed and why, of one who is critical of all the suggestions and who is the ultimate owner of all the outputs.
The GPS principle operationalizes the four dimensions of AI literacy identified by EDUCAUSE: technical understanding, evaluative skills, practical application, and ethical considerations, which do not require the mastery of AI theory prior to its application in the framework.10
The course is designed in terms of five ethical principles of applying AI to the workplace: transparency, fact-checking, finding bias, holding oneself responsible, and privacy protection. These are not moral categories that are abstract. They are applied to provide training in the form of a scenario. What would transparency be like when AI writes your cover letter? What does fact-checking mean when the AI is listing qualifications not listed in the job posting? What does being able to detect bias entail when AI continues to undervalue much of the career experience of a student? The competency framework created by Chee and others offers research grounding to these differences, such as the recorded variations in AI literacy requirements among groups of learners. It validates that the ethical aspect of AI use cannot be neatly divided from the practical one.11
The least impactful design decision in the course is also the easiest: Each big deliverable will have a structured self-reflection of 200 to 300 words. AI cannot recreate a reflection of a networking chat that occurred in person, or a career panel that the student attended, or a professional crisis they had to resolve independently. The reflection is the section of the assignment that only the student can write.
The eight-session arc includes:
- AI ethics and an introductory AI quiz,
- job search strategy,
- day-in-the-life career exploration,
- professionalism and personal branding,
- LinkedIn profile development,
- applicant tracking systems,
- resume tailoring and optimization, and
- a networking activity with a required reflection.
Students are not just learning how to be career-ready in this course. They are learning the paradigm of critically assessing their relationship with AI in any professional setting they will find themselves in the future.
The GPS Principle in Practice
The GPS principle is strong because it serves as a framing metaphor, but its practical worth for our GENS 401 course is in the question of accountability it creates in any assignment: Can you have someone ask you about this document during an interview?
The job-search strategy session involves the students applying AI to perform Boolean searches and set up target company alerts, and then bring the results back and describe in their own words what opportunities they decided to pursue and why. AI did the search. The student determines which job to move forward with for the assignment.
In the LinkedIn profile development session, students have an AI point out the differences between their current profile and those of people in their desired profession, and compose their own text—as accurate and relevant—to fill the gaps in language that sounds natural to them. AI named what was missing. The student provided what was real.
In the resume tailoring session, students copy a job posting into an artificial intelligence application, which will then suggest the most commonly used keywords. The students may update their own resume with the best-matching keywords in their own words and based on their own experience. AI flagged the gap. The student authentically addressed it.
The human step is not optional in any of the cases. It is imposed structurally by the requirement of reflection. It is applicable outside of the class itself because of the accountability question. Ideally, students will remember what they learned and apply it moving forward in their career journey and professional development.
What the Evidence Revealed: The Authenticity Gap Closes With Structure
I want to be clear about what I share next. These are practitioner observations across two semesters of systematic delivery, not findings from a randomized controlled study, a blinded assessment design, or a large quantitative dataset. I am presenting them with that framing made explicit, because overclaiming in the practitioner literature does the field a disservice.
With that said, what I observed, consistently, across both semesters of GENS 401 was that students who engaged seriously with the GPS principle and the structured reflection requirements produced application materials that were qualitatively stronger and more coherent with how those same students described themselves in self-reflections.
I observed this same pattern in career counseling appointments where I applied the same framework to students who visited me for class resume review or mock interview assignments. The rubrics we used in both situations measured alignment between written materials and verbal self-presentation. The alignment was meaningfully better among students who completed the assignments with authentic engagement. That consistency across two delivery cycles is enough to warrant continued development and, I would argue, enough to warrant the attention of the broader career services field. There was also an additional pattern revealed in the data that we did not design for, nor did we set out to study.
A Finding About Equity That Deserves Careful Attention
We did not design GENS 401 as an equity intervention. We did not set out to study first-generation or neurodivergent students as a distinct population. What I observed across two semesters of the course and in my daily career advising appointments, which regularly include both groups, was not what I anticipated.
I expected confusion about what AI use was permissible. I expected some frustration that AI was acceptable in career readiness when it was not in academic coursework. I expected resistance. What I got, from first-generation and neurodivergent students in particular, was relief.
That reaction is understandable given the barriers that have been documented in career readiness by these students. First-generation college students continually lag behind their continuing-generation counterparts in terms of networking activity, career management, and professional communication skills, and only 16% turn to career centers during their first year.12 The barriers are not attitudinal. They are institutional: the lack of entry to professional networks, the lack of understanding regarding the unwritten professional norms, and a lack of informal mentorship and observation on which continuing-generation students feed unknowingly. Nester, writing for NACE, identifies AI as a potential equalizer for first-generation students, a way to bridge gaps in professional knowledge and networking access that these students have historically had less exposure to.13
For neurodivergent students, those challenges compound. A scoping review by Bölte and colleagues documented significant barriers to career readiness in neurodivergent individuals, including self-advocacy in professional environments, building career identities within communication systems designed around neurotypical norms, and navigating the ambiguous, unspoken social expectations that most professional development resources take for granted.14 A qualitative case study of university students with high autistic traits in Malaysia identified structured experiences and clear professional expectations as central to the career confidence students described.15
Without a designated structure, the students I advise in either category are likely to revert to outputs of AI that seem sophisticated but generic, the worst of all possible results to achieve in a group that must stand out and not blend in. The GPS principle provided them with clarity regarding a process that had been hidden. The observed change in student confidence over the two semesters was not minor. Although I am not able to measure it formally as a researcher would, as a career coordinator, I can confirm that what I observed was regular, it existed, and it went in one direction. What I found was that structured AI fluency does not just equip students to handle professional expectations. It balances access in a way that matters most for students who have been historically disadvantaged by unspoken professional norms.
What Career Services Can Do Right Now
Career centers do not require a complete overhaul of their curriculum to begin changing how students engage with AI. The three interventions that follow are scaled by institutional investment, ranging in intensity from one question in an advising appointment to a full course structure. All three can be implemented using available resources. The only thing they have in common is one core principle: the student remains the author of their own career narrative.
Intervention 1—the advising question: The simplest intervention involves simply a question—a single question, spoken during all advising sessions using application materials: Can you tell me how you applied AI in this case and where your own voice was used?
That question puts the whole advising conversation into another context, without accusations or confrontations. It provides an opportunity for the student to be supported in understanding that they are the one who is in charge of the content that has been produced under their name. It introduces a natural dialogue answering the question of the distinction between AI as a thinking partner and AI as a ghostwriter. And it starts developing the metacognitive practice of being an author—the cognition that professional writing is not merely production. They are professional reflections of an individual. They follow you. They influence your reputation and credibility even after the time you have clicked apply.
As soon as students understand it, they will be far more critical of the current approach they take to AI use and will be conscious of using their authentic voice on the job application materials to make them shine. That is what employers want to see, and that is what helps students stand out as top applicants in a pool of generic candidates.
Intervention 2—GPS workshop: For practitioners desiring a structure but not a full course, the GPS principle can be directly applied to a 60-to-90-minute standalone workshop. It may be provided as a standalone event, as a series related to the recruiting season or career fairs, or as a complement to existing resume review programs.
In the workshop, the navigation tool and driver framing are presented, the students are guided through the side-by-side analysis of responsible and irresponsible AI applications in real-life job search activities, and the responsibility question is introduced as a question that can serve as a portable standard of self-assessment. A short-structured reflection accompanies each practice activity. This is crucial in a landscape where faculty confidence in their ability to teach students how to use AI ethically is only at 32%, a professional development gap that cannot be solely absorbed by career services.16 That gap can be bridged, without institutional policy or curriculum redesign, with a workshop that provides that named, memorable framework that the students can use on their own.
Intervention 3—full course model: The GENS 401 course runs eight weeks, one credit hour, fully online and asynchronous, with the five ethical principles taught through career-application scenarios, formal reflections embedded in every major deliverable, and the accountability question used as a recurring self-assessment standard throughout every session.
Two principles guide implementation. First, introduce the authenticity gap during the first week of any course in which the students are using AI to do any writing; this ensures they aware of the professional stakes involved from the very beginning. Second, build structured reflection into every assignment involving the use of AI, so that the human step cannot be skipped, and verify it through an instructor review that looks for alignment between the student’s written work and verbal self-presentation.
The framework is intended to be customized, rather than copied. The sequence of assignment tasks is not quite as important as the concept: AI is used for analysis, evaluation, and identification, but not meaning-making, which is done by the student.
Note: For institutions with the capacity to offer a credit-bearing career readiness course, the full GPS principle framework, as implemented in GENS 401, is available for free as an adaptable model from the author on request.
The Challenge Predates the Technology
The hiring ecosystem is facing a crisis of authenticity that neither recruiters nor students created on their own, and that neither can solve alone. Recruiters cannot easily tell if an application in front of them was submitted by an AI bot or a human being. Students are not able to stand out in a sea of applicants using AI’s homogenized output. The career services function between them has never been more needed, more necessary, or more contested.
We are not here to help students compete for jobs. We are here to insist that the person behind the application matters. That their authentic voice is a professional asset. Being able to stand in a room and defend their professional documents under genuine questioning is not optional. It is the entire point.
Teaching students to be accountable, critical, and transparent in their AI use is not a response to a technology problem. It is a response to a human development challenge that predates these tools and will outlast whatever platforms come next. Only 32% of faculty globally feel confident supporting students in ethical AI use..17 Career services professionals cannot afford to be part of that statistic. The five ethical principles, the GPS principle, the structured reflection, and the accountability question are not AI-specific interventions. They are the crucial career development interventions that meet students where the work has always been—in the creation of a professional identity that can be articulated, defended, and owned with confidence.
Endnotes
1 Greenhouse. (2025). 2025 AI in Hiring Report: An AI Trust Crisis. Retrieved from www.greenhouse.com/newsroom/an-ai-trust-crisis-70-of-hiring-managers-trust-ai-to-make-faster-and-better-hiring-decisions-only-8-of-job-seekers-call-it-fair.
2 Ibid.
3 Ibid.
4 Patterson, D. (2025, September 16). The Hiring Hoax: What 3,000 Managers Revealed About AI & Identity Fraud in 2025. Checkr. Retrieved from https://checkr.com/resources/articles/hiring-hoax-manager-survey-2025.
5 Clemo, F. (2026, January 19). Deepfakes and AI-enabled Impersonation Rank Among Top Recruitment Threats, Research Reveals. People Management. Retrieved from www.peoplemanagement.co.uk/article/1945557/deepfakes-ai-enabled-impersonation-rank-among-top-recruitment-threats-research-reveals.
6 Paoli, N. (2025, November 18). Trust Is at an All-time Low for Both Job Seekers and Recruiters. Fortune. Retrieved from https://fortune.com/2025/11/18/hiring-job-seekers-recruiters-talent-acquisition-ai-doom-loop-application-technology/
7 McKinsey & Company. (2025, March 12). The State of AI: How Organizations Are Rewiring to Capture Value. Retrieved from www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai-how-organizations-are-rewiring-to-capture-value
8 Purcell, K. (2025, July 14). 2025 Applicant Tracking System (ATS) Usage Report: Key Shifts and Strategies for Job Seekers. Jobscan. Retrieved from www.jobscan.co/blog/fortune-500-use-applicant-tracking-systems/.
9 Gatta, M. (2026, April 30). Benchmarks: AI Integration in Career Centers Is Growing. NACE. Retrieved from https://naceweb.org/career-development/trends-and-predictions/benchmarks-ai-integration-in-career-centers-is-growing.
10 Kassorla, M., Georgieva, M., & Papini, A. (2024, October 17). AI Literacy in Teaching and Learning: A Durable Framework for Higher Education. EDUCAUSE. Retrieved from www.educause.edu/content/2024/ai-literacy-in-teaching-and-learning/executive-summary.
11 Chee, H., Ahn, S., & Lee, J. (2025). A competency framework for AI literacy: Variations by Different Learner Groups and an Implied Learning Pathway. British Journal of Educational Technology, 56(5), 2146–2182. https://doi.org/10.1111/bjet.13556.
12 Montoya, D. (2022, June 1). Serving Our First-generation University Students. Career Convergence Web Magazine. Retrieved from www.ncda.org/aws/NCDA/pt/sd/news_article/441488/_PARENT/CC_layout_details/false.
13 Nester, H. (2024, August 22). How First-generation College Students Can Use AI to Level the Job-search Playing Field. Summer 2024 NACE Journal. Retrieved from www.naceweb.org/career-development/best-practices/how-first-generation-college-students-can-use-ai-to-level-the-job-search-playing-field.
14 Bölte, S., Carpini, J. A., Black, M. H., Toomingas, A., Jansson, F., Marschik, P. B., Girdler, S., & Jonsson, M. (2025). Career Guidance and Employment Issues for Neurodivergent Individuals: A Scoping Review and Stakeholder Consultation. Human Resource Management, 64(1), 201–227. https://doi.org/10.1002/hrm.22259.
15 Zainal, M. S. (2025). Empowering Students With High Autistic Traits: Advancing Career Preparedness and Inclusive Development in Universities. Multidisciplinary Science Journal. https://malque.pub/ojs/index.php/msj/article/view/11507.
16 Rettler-Pagel, T., & Grovergrys, K. (2025, May 8). Career-ready in the Age of AI. Community College Daily. Retrieved from www.ccdaily.com/2025/05/career-ready-in-the-age-of-ai/.
17 Ibid.
