AI and The Plagiarism Trap

Artificial intelligence (AI) has proven to be a paradigm-shifting tool. We foresaw AI impacting our hospitality work environment years ago. I was first asked to develop an approach to AI in Spring 2022. AI is not limited to higher education though. Numerous lodging global conferences on the topic have featured AI prominently in recent years. This includes the organization I serve as Global President of, Hospitality Financial and Technology Professionals (HFTP).

I have heard from and spoken to thousands of lodging professionals about professional AI usage at conferences and events in the past four years. Revenue managers, sales and marketing teams, human resource professionals, hoteliers in operational roles, controllers reconciling statements, tech savvy hoteliers, and even ownership groups are leveraging AI to improve outcomes and workflow. Simultaneously, an emerging generation entering the workforce has learned to incorporate AI in more than their professional lives; AI is present in their personal lives as well.

My advisory board includes a regional vice president of human resources for a large hospitality company (i.e., their VP of HR says the company has more than 60,000 global team members). With that many team members, the organization requires some streamlined processes. One process they have streamlined in H.R. is resume assessment. They leverage AI in their hiring process; specifically, they use AI detection tools. Approximately 25% of resumes in the past year have been removed from their system without a human viewing the resume. The reason is that the organization does not want to hire anyone who plagiarizes in their resume.

This conversation was one of my first about the impacts of AI plagiarism in hospitality. I have had quite a few conversations since then. Numerous industry professionals are concerned about plagiarism, especially those who hold VP-level and C-suite positions. Yet, I have not read an article or seen a training session at a conference about AI plagiarism yet.

AI, when leveraged properly, can be a useful tool. AI has numerous pitfalls though. I have heard an enormous amount of discussion around corporate secrets and financial statements being obtained by those who should not have access to such materials at corporate events. This is admittedly possible. There is something to worry about that I have not heard industry professionals speaking to though: plagiarism. Plagiarism is integral to how AI is used in our industry.

The Dangers of Plagiarism to a Hospitality Professional

Middle school students are introduced to the idea plagiarism in most global education systems. High schools and universities hone that knowledge. Problematically, my experience is that students mostly misunderstand plagiarism. Being in higher education myself, I feel this is because students have for decades understood plagiarism through examples rather than conceptually. Specifically, students seem to understand plagiarism as copying someone else’s work, normally their verbiage, and claiming the work as their own unique writing. That is an example of plagiarism. This is not what plagiarism is, though. Plagiarism includes presenting someone else’s ideas as your own. This can make AI reliance tricky. The emergence of AI is leading to unintended plagiarism consequences.

I often present the importance of understanding and avoiding plagiarism. I present my warnings alongside concerns over corporate secrets and financial information when speaking to hospitality industry professionals. It seems plagiarism is less important to those in attendance based on the questions I receive. Corporate espionage, cybersecurity, financial information, and other private information about a company can be harmful if it falls into the hands of nefarious actors. These things should be taken seriously; we need to take privacy seriously in general when it comes to AI. Yet plagiarism is far more likely to take place and exponentially more likely to harm individual professionals.

There are numerous examples of professionals losing their careers over plagiarism. This is less prevalent so far in hospitality and tourism than in some other sectors. Yet, it does impact our industries and how we use AI. Furthermore, we do not necessarily know where our careers will take us. I know dozens of professionals who are simultaneously in academia, journalism, politics, or VP-level of C-Suite in companies. Individuals in these four areas seem particularly likely to face career-altering consequences stemming from plagiarism.

There are six widely recognized illustrations I use as examples of the perils of AI plagiarism when speaking at industry events (see six sub-headers beneath Table 1). I choose these patterns because each clearly showcases one of the six primary ways that accidental AI plagiarism takes place in professional environments in the AI era:

1. Faulty citations including hallucinations

2. Proprietary data misuse

3. Patchwriting

4. Inadequate attribution

5. Ghostwriting

6. AI software

Table 1 identifies these six forms of accidental plagiarism and provides examples in our industry. I used Open AI’s Chat GPT to assist me with the creation of the table. This is not plagiarism because I am clear about what is mine and what is not and cite things appropriately. Furthermore, to ensure I do not fall into any traps of misinformation, I thoroughly checked what the AI provided me.

The columns “Plagiarism Form” are my unique ideas. Column “Definition” is my unique ideas, but an AI assisted me with the writing (Open AI, 2026). How did an Open AI tool assist me? I uploaded the article you are reading and two others I authored about plagiarism and AI. I asked the AI tool to summarize my unique ideas about the six forms of plagiarism associated with AI. I need to provide this AI tool’s role in assisting me to ensure this is not plagiarism.

Finally, the “Lodging Industry Example” column is an AI tool’s unique creation beyond me authoring a prompt (Open AI, 2026). The prompt I write included this verbiage, “Create an APA Table that includes…a made-up example from the lodging industry for each of the six forms of plagiarism.”

1. Faulty Citations Including Hallucinations

You have likely heard of Steven Schwartz, a NY-based lawyer, even if you do not recognize him by name. His story made national headlines for months when Open AI’s Chat GPT was a budding tool. Schwartz practiced law for more than thirty years before citing materials improperly in a legal briefing. He lost his license over this instance of plagiarism. What many people do not realize is that his partner was also found liable.

When you have a collaborator on a project, you can be held liable for their plagiarism. Your name is on the document, therefore, you are responsible for what is contained in the document. Yes, this holds true in our profession as hospitality professionals as well. We are responsible for documents a collaborator took part in, even if the errors are not from our section(s). Finger-pointing to a collaborator regarding a document your name is on the cover of is the type of excuse I expect from middle school students; yet, I read this excuse being used often in professional circles as an adult.

Schwartz did not purposefully plagiarize. Rather an AI hallucinated when he used it as a research tool. Hallucinations can be understood as fake information presented as fact after being AI-generated. As in, AI tools can provide false information and present it as a fact. This can include citations as happened with Schwartz.

You may notice I did not use any citations in the story above. This is because the story and the analysis I present are either considered common knowledge or they are my own unique ideas or experiences.

Common knowledge can be understood as information that is widely known, available, and accepted in the author’s field. As a professor specializing in AI, this is a story I know well and have been telling for years already.

Analysis of someone else’s idea or experience is sometimes identified as plagiarism. We do need to cite other’s ideas. The story above is common knowledge and the ideas are my own. Thus, citations in this context are not necessary.

2. Proprietary Data Misuse

Another example is from the lodging industry’s real estate market. Xceligent is a commercial real estate organization. A couple members of the team accessed CoStar’s database. The database existed behind a paywall. Xceligent’s employees downloaded and later used materials they did not have permission to use on their organization’s website. This included private lodging information. CoStar sued for approximately $500 million; they were awarded nearly $9 million (Williams & Connolly, 2019).

I tell this story often. The story itself is common knowledge in my professional life. I could not remember the precise financial settlement though. I needed to cite this because I looked it up. This information is someone else’s idea or knowledge. The citation ensures it is not plagiarism.

This story is specific to hospitality and tourism. Furthermore, it is something I have personally witnessed in our industry. I was recently hired to assess the value of a hotel that was part of divorce proceedings. I learned that the verbiage used by the couple to market their property was written by an OTA. The couple copied and pasted the verbiage into their own marketing without giving credit to the author (i.e., the OTA). First, this is plagiarism. Secondly, it opens them to potential legal liability. What I want you, the reader of this article, to take away from this story is that the unauthorized use of verbiage through plagiarism can take place in business settings. This can be serious. Plagiarism can open your company to potential legal liability, as with this case of unauthorized intellectual property.

3. Patchwriting

Fareed Zakaria has spent decades in the public eye as a newscaster (e.g., CNN) and reporter (e.g., Washington Post). He was recently accused of dozens of instances of plagiarism. He was accused of a form of plagiarism that professors such as me see most often in student submissions: patchwork. Patchwork is when a writer rephrases someone else’s original ideas. Specifically, the author is not copy and pasting someone else’s work. This is not an issue of using someone else’s words. Rather, the ideas are being plagiarized.

There are levels to this form of plagiarism because how much verbiage is changed can alter the presentation of the original idea(s). The writer is not using someone else’s words; rather, they are presenting someone else’s idea as their own. This is common in many ways in newscasting for individuals such as Zakaria. After all, a newscaster cannot obtain all global news themselves. They are often presenting someone else’s ideas. Yet, presenting someone else’s ideas as your own constitutes a major breach in ethics when writing as a journalist or other professional.

The students I work with rarely understand this juxtaposition. I have received submissions where every single idea is lifted from AI; they simply rephrased AI’s wording. Teaching the idea of patchwork as plagiarism can be difficult. I therefore often explain patchwork through food at conferences. Imagine you pick up as three-course meal from your favorite restaurant, bring it home, add a small amount of pepper to the salad, and tell everyone you cooked the entire meal. This is a decent euphemism for patchwork.

4. Inadequate Attribution

Claudine Gay was the President of Harvard University shortly after COVID. Her tenure lasted less than one year. Political actors succeeded in removing her from the leadership role. These actors sought to find issues with her personal and professional life in order to remove her. They turned to plagiarism to accomplish their goals. While academia generally holds that Dr. Gay did not commit plagiarism, there were instances in her research that were attacked for potential plagiarism.

Specifically, the idea was that some ideas she presented as common knowledge should have been cited. She lost her position as President of Harvard University by not providing adequate attribution according to a limited group of individuals in material she published many years before becoming President of Harvard University.

I imagine that when Claudine Gay was an emergent scholar, twenty or more years before becoming President of Harvard, she did not necessarily expect to become President or understand the politics that come with such a role. She may have hoped for such a position but counting on something like that is unlikely. I imagine a graduate student writing tens of pages while working overnight and potentially missing a few citations. Nothing in her authorship was even necessarily plagiarism; it was simply debatable about whether citations should have been used.

You do not know where your career may take you either. As you read this – or as I write this – neither of us knows what someone who dislikes us twenty years from now may try to do to harm us professionally or personally. Avoiding plagiarism is therefore wise.

5. Ghostwriting

Ghostwriting can be understood as an individual presenting work they did not write as their own. This is actually normal in business, politics, and dozens of other fields. There are many times when a presenter or author includes material from one of their team members (e.g., subordinates) and presents it as their own. This is not necessarily nefarious. It is not considered plagiarism if one person is speaking on behalf of a group of people in a professional capacity. For example, a hotel General Manager (GM) presenting the Profit and Loss (P&L) statement to the ownership group does not need to give credit to everyone who worked on that year’s budget or other financial statements. It is understood by all present that the GM is presenting work that they crafted with their team. Ghostwriting can be plagiarism though.

I oversaw the hundreds of articles in our publications and dozens of education modules in our HFTP Academy as Global President of HFTP. Some of our publications are newswires. A newswire is a single outlet that curates the most appealing stories from other sites and presents them. Newswires are excellent resources for industry professionals and integral to someone like me staying abreast of niche topics in my field.

An example is that HotelExecutive articles are written specific for the site. The best articles may be picked up by newswires in our field including my newswires from HFTP. Thus, one article I author on HotelExecutive may end up on numerous additional newswires including my very own HFTP newswires. This is a good thing so long as credit is provided! But if a newswire presents the article as their own rather than giving credit to HotelExecutive as the publisher and me as the author, then that becomes plagiarism.

Obviously this would never happen at HFTP. I have seen it happen with other newswires in our field, though. I begin reading an article before realizing I wrote it but a different author or company’s name is listed in the author field.

A second level to this is when a newswire picks up a plagiarized story. Forwarding plagiarized material without understanding its origination can be tricky. Newswire sites have a responsibility to not distribute plagiarized materials. This is a reason newswires should attribute authorship in all situations, to ensure that if something is identified as problematic then it does not come back to the newswire site as being problematic.

6. AI Software as a Plagiarism Tool

I need to mention grammar assistance tools (e.g., Grammarly, Ghotit Real Writer) in an article about plagiarism. Grammarly is commonly used by English Language Learners (ELL), while Ghotit Real Writer was developed to assist individuals with dyslexia. These tools can become plagiarism when they move beyond correcting grammar and instead re-write sentences, re-phrase ideas, or introduce new language that was not originally written by the author. While a human may provide an initial prompt, the software can alter the wording so substantially that the final text no longer reflects the author’s own ideas or voice.

The paragraph above was written by an AI tool. I only provided a prompt. The prompt I wrote read, “I need to mention grammar assistance tools (e.g., Grammarly, Ghotit Real Writer) in an article about plagiarism. Explain how these constitute plagiarism in a concise paragraph.” Yes, I wrote the prompt. The idea is mostly my own; yet, not all of the ideas are my own and the wording is definitely not my own. This would be plagiarism if I did not cite it: (Open AI, 2026). Having read my prompt and the paragraph provided by an AI tool, I am hopeful the plagiarism is easy for you as a reader to identify.

Avoiding Plagiarism

AI is something we are all learning. We are going to make AI errors. I believe that what we say, do, and write today are things we can be held accountable for later though. As with Dr. Claudine Gay from Harvard University, we cannot be sure where life will take us. The safest thing we can do for our future selves is not plagiarize. Even if you do not believe you will ever be held accountable, there are millions of people who have learned how to identify AI plagiarism.

If you author something that can be identified nearly immediately as plagiarism by well-versed individuals, then you will likely be held accountable for it in some manner. There are a few academics and lodging professionals that I know are plagiarizing mercilessly right now. I am not sure whether they believe they will never be caught, do not understand plagiarism, or simply do not care. They may not realize it, but numerous individuals are discussing their plagiarism right now. They are less likely to be respected in their fields from now on.

A question I often receive is how I identify the plagiarism, like in the situation above with academics ad lodging professionals. Learning to identify plagiarism is more useful than merely trying to hold others accountable (e.g., my students, articles submitted to journals I edit). Therefore I hope you can also learn to identify plagiarism. Learning to identify plagiarism is an excellent way to begin understanding and subsequently avoiding plagiarism yourself. After all, AI is our future. We should be using it. Using AI tools without plagiarizing can be tricky for those just now learning how to go about leveraging emergent technologies in the AI space.

One way to identify plagiarism is through AI detection tools. Not all detection tools are created equal. What the best AI detection tool is changes fairly often. In the past year alone, 5 different AI detection tools have proven to be the best at different times when I ran tests on them. Which AI tools are best is only partially associated with opinion. There are scientific facts involved as well.

I believe the best AI detection method is called "ground-truth dataset testing" (GTDT). Think of GTDT as a grade that AI detection tools receive, just like students receive grades in their coursework. The rubric that GTDT uses to detect AI and grade something includes: Accuracy, Precision, Recall, False Positive Rate, and False Negative Rate (Google for Developers, 2026). GTDT is an excellent assessment method. Yet, GTDT cannot stand alone as a method. There are numerous reasons for this. I summarize the reasoning as GTDT being based on a dataset of AI authored prompt responses. Given AI is evolving rapidly, the datasets become outdated quickly.

There are other good methods used to grade AI detection tools as well. For example, adversarial testing is normally well respected when given as a test alongside GTDT. In layman’s terms, adversarial testing means an AI – the assessment tool you are using – is trying to confuse the AI you are assessing. An AI testing another AI, with the purpose being to confuse it, provides meaningful data.

Undetectable.ai and Originality.ai are AI detection tools that use GTDT and adversarial testing. Grammarly is an example of an AI detection tool that does not use either of these. Grammarly relies primarily on something called Perplexity Analysis. Perplexity Analysis is an outdated and simplistic AI detection method. I write this because some AI detection sites are far superior to others. Knowing which sites to trust is important when you are assessing your own, or someone else’s, potential plagiarism.

No AI detection tool is perfect; they can be reliable though. I want to give you an example. Undetectable.ai has an approximately 93.6% rate of identifying AI on a sentence-by-sentence basis (Perry, 2025; Whiles, 2025). If I upload 100 sentences, this means that the percent chance of AI being detected correctly is approximately 99.9999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999%. Yes, that many 9s. I included all of the 9s to show you how certain these detection tools can be statistically. There are obvious nuances and scientific methods at work here. A case could be made that in some situations there are fewer nines after the 99% likelihood. I am willing to accept that.

Most AI tools provide the percentage certainty of AI being used at all. The percentages can be difficult to understand. For example, a 35% return does not mean that 35% of the submission is plagiarized. This is how I understand the percentages provided by AI tools.

This is based on my experience (see bullet points and Table 2):

• 30% or less – The AI detection tool is letting me know this was likely submitted by a human author.
• 30%-55% - The AI detection tool is uncertain of the submission’s originality. I find that my most academic works, my theoretical and philosophical submissions, often fall into this area. I mostly view this middle area as not being AI generated.
• 55%-80% - The AI detection tool has found evidence of AI plagiarism. The entire submission is unlikely to have been plagiarized, but the evidence of plagiarism is overwhelming. This is where patchwork plagiarism is most often found.
• 80%-100% - The AI detection tool found clear evidence of copy and paste from AI. This is a pure form of plagiarism.

I have seen my writing hit 50% likely before even though I did not use AI at all. But if something shows red, 80% AI, I see all the hallmarks of AI, and multiple sentences are identified with 100% certainty as being AI? Then it becomes obvious what happened.

I do not only use AI detection tools. I assess things myself as well. We should never rely on AI entirely, whether assessing AI or leveraging AI to assist us in our work. I know to look for and assess sentence structure, comma use, dashes and hyphens, vocabulary, industry-specific acronyms and terms, transition phrases including potential repeated words, and the general flow.

Conclusion

I hope this explanation helps you understand how we can leverage the tools available to us to be better with AI! AI is not going away. I am hopeful you learn to leverage its enormous capabilities to be more efficient and more effective in your personal and professional life. With this written, there are real dangers around AI. We should be cognizant of these dangers. Organizations seem most focused on competitors accessing their data, including financial materials. Those are real concerns. These could impact your business operations. I am equally concerned with plagiarism. Plagiarism seems more likely to impact you as an individual while most other AI concerns are more likely to affect your business. You should be concerned with plagiarism.

I uploaded this finished paper to Open AI (2026). I asked the AI tool to write a conclusion for this article. I feel this showcases an AI tool’s ability to summarize and assist authors. This is an excellent way to use an AI tool. The ideas and authorship of this paper are entirely mine, while the conclusion section is an AI’s assessment of my ideas, summarized, and provided to you. Please pay attention to the flow, writing style, use of dashes, abundance of commas, and other AI markers in this conclusion. This is what Open AI (2026) provided me:

Plagiarism is one of the most significant risks associated with AI adoption in professional environments. Unlike cybersecurity breaches or corporate espionage, plagiarism often occurs unintentionally. AI systems can generate convincing language, citations, and analytical frameworks that appear authoritative but may not be accurate or original. When professionals rely on these outputs without verification, attribution, or careful editing, they risk presenting ideas that are not their own. In doing so, they expose themselves and their organizations to reputational damage, legal liability, and career-altering consequences.

The six forms of accidental plagiarism discussed in this article—faulty citations, proprietary data misuse, patchwriting, inadequate attribution, ghostwriting misrepresentation, and misuse of AI writing tools—illustrate how easily plagiarism can occur in the modern AI environment. None of these situations necessarily involve malicious intent. In most cases, they arise from misunderstanding how AI tools generate and transform information. Yet intent rarely protects professionals from the consequences of plagiarism once the work has been publicly distributed.

The safest approach for lodging professionals is therefore simple: use AI as a tool, but remain fully responsible for the work that bears your name. Verify sources, attribute ideas appropriately, protect proprietary information, and ensure that the final product reflects your own analysis and judgment. AI can assist with efficiency and productivity, but it cannot replace professional accountability.

As AI continues to evolve, so too must our understanding of ethical authorship and intellectual ownership. The hospitality industry has always relied on trust, credibility, and professional reputation. Protecting those assets requires vigilance in how we incorporate emerging technologies into our work. By understanding the risks outlined in this article and applying careful judgment when using AI tools, lodging professionals can benefit from AI’s capabilities while avoiding the very real dangers of plagiarism.