20 types of plagiarism and academic misconduct
Introduction: New academic integrity challenges
When ChatGPT, a revolutionary generative artificial intelligence (GenAI) tool, was released to the public in November 2022 (OpenAI, 2022), it sparked a wave of innovation throughout the AI landscape. Its widespread popularity, along with the rapid emergence of other AI writing and research tools, has dramatically transformed the educational sector. As a result, the notions of academic misconduct and plagiarism are becoming more varied and complex, with new and sophisticated forms arising from the misuse of AI tools.
Nearly 7,000 higher education students in the UK were caught cheating using ChatGPT and other AI tools in the 2023-24 academic year (Goodier, 2025). The investigation by The Guardian reveals that traditional forms of plagiarism are declining, while AI-related integrity violations have surged significantly in recent years (Goodier, 2025). This issue has extended beyond the classroom and infiltrated published scholarship. In November 2025, a doctoral thesis authored by a PhD student in Hong Kong and published in an academic journal was found to contain fictitious or non-existent AI-generated citations and references (Ho, 2025; Hung, 2025). This incident has sparked widespread concern and discussion.
In this transformative AI era, understanding what academic misconduct and plagiarism mean today has never been more crucial. This article examines their contemporary definitions and various forms, drawing information from the websites of most Russell Group universities and other reputable organisations. Whether you are writing essays or doing research, staying informed is the first and most essential step towards achieving academic integrity. It is a must-read for anyone seeking to avoid academic misconduct and plagiarism and enhance their writing in today’s academic landscape.
Defining academic misconduct and plagiarism in the age of AI
Academic misconduct refers to any actions, attempted actions or omissions that may result in a student obtaining an unfair academic advantage in an examination or assessment (UCL, n.d.-a; University of Warwick, n.d.). This can also include actions that assist someone else in obtaining an unfair advantage, or activities that are likely to undermine the integrity essential to scholarship and research (University of Warwick, n.d.). Common types of academic misconduct include plagiarism, self-plagiarism (which some universities classify as a type of plagiarism), collusion, contract cheating, falsification, fabrication, impersonation in any form of assessment, examination room misconduct, and online assessment misconduct, among others (UCL, n.d.-a; University of Nottingham, 2025a).
Unlike academic misconduct, which encompasses a broader range of dishonest or unethical behaviours, plagiarism is the most recognised type of academic misconduct and has various specific forms. Plagiarism is defined as “the process or practice of using another person’s ideas or work and pretending that it is your own” (Cambridge Dictionary, n.d.). It can arise from all types of sources and all forms of media–not just written texts. This includes graphical materials, music, computer codes, videos, audios, calculations and applies to both published (e.g., journals, books, websites) and unpublished materials (e.g., lecture notes, other students’ work), whether in manuscript, printed or electronic form (University of Oxford, n.d.; University of Nottingham, 2025a). Therefore, content generated wholly or partly using AI also falls under this definition (University of Oxford, n.d.). Common types of plagiarism are direct plagiarism, complete plagiarism, paraphrasing plagiarism, accidental plagiarism, source-based plagiarism, patchwork plagiarism, and so forth (Kramer, 2022).
In the current AI era, some traditional forms of plagiarism and academic misconduct, such as source-based plagiarism and contract cheating, have become more complex, while introducing new forms like AI misuse. To maintain academic integrity, students should be the authors of their own work; using text generated or paraphrased by AI undermines this integrity and fails to meet the originality requirement (University of Cambridge, n.d.-a). Since AI tools are incapable of producing original ideas and taking responsibility for their outputs, they cannot be regarded as authors (University of Glasgow, 2024).
20 types of plagiarism and academic misconduct: Traditional and emerging forms
The terms “plagiarism” and “academic misconduct” are defined differently across reputable UK universities, with some institutions using them interchangeably. These differing definitions lead to variations in their scopes and recognised types. For this reason, this section discusses both types together rather than separately, with a focus on traditional and emerging forms of misconduct in academic writing through an exhaustive list. Understanding these variations is vital for preventing violations of academic integrity in today’s educational environment.
Equally important, it is often encouraged to conduct a plagiarism check before submission (UCL, n.d.-a). Some UK universities, including UCL, offer their students and staff the opportunity to check their work confidentially using the Turnitin Similarity tool (UCL, n.d.-b). Such checking is considered responsible because it helps you to identify flagged areas, improve your citation and referencing skills, and make necessary manual edits using your own judgment, knowledge, and writing skills.
1. Direct plagiarism
Direct plagiarism, also referred to as word-for-word plagiarism, verbatim plagiarism, or clone plagiarism, involves deliberately reusing another’s language, data, and/or illustrations exactly as they appear, without using either quotation marks or indentation, or providing full referencing of the sources cited (University of Cambridge, n.d.-a; University of Oxford, n.d.).
2. Complete plagiarism
Complete plagiarism, one of the most severe forms of academic offence, occurs when an individual copies an entire work from another source without proper acknowledgement and falsely claims it as their own (Kramer, 2022). Whilst direct plagiarism involves copying specific sentences or paragraphs without proper quotation marks or citation of the source, complete plagiarism goes further by submitting the entire paper as if it were original (Kramer, 2022).
3. Paraphrasing plagiarism or modifying
Using another person’s ideas by simply changing a few words, altering the order of presentation, or closely following the structure of their argument without proper acknowledgement (University of Oxford, n.d.; University of Warwick, n.d.).
4. Borrowing/using ideas or concepts
Using ideas or concepts taken from others without acknowledging the source, even if you express them in your own words, violates academic integrity (University of Cambridge, n.d.-a; University of Warwick, n.d.).
5. Cutting and pasting from the Internet without clear acknowledgement
Cutting and pasting content from the Internet without crediting the original sources is considered plagiarism; any Information taken online must be cited and listed in the bibliography (University of Oxford, n.d.).
6. Accidental plagiarism
Accidental plagiarism, also known as inadvertent or unintentional plagiarism, occurs when someone plagiarises without intending to do so. This type of plagiarism is not always deliberate. Students may forget to quote or cite a source, be genuinely uncertain about how to reference it correctly, or fail to realise that they are paraphrasing another’s work (Kramer, 2022; Turnitin, n.d.). However, a lack of intent does not free you from academic penalties. It is your responsibility to learn your discipline’s referencing conventions, take clear notes, and avoid closely paraphrasing (University of Oxford, n.d.).
7. Failure to acknowledge assistance
It is vital to explicitly acknowledge the support and advice you received from fellow students, laboratory technicians, and other external sources that have contributed to your work, including guidance that led to significant changes in content or approach (University of Oxford, n.d.). In cases of joint projects or collaborative research, it is essential to explicitly identify which parts of the work were completed by others (University of Cambridge, n.d.-a).
Nevertheless, there is no need to acknowledge help from your lecturer or supervisor, nor assistance of proofreaders (University of Oxford, n.d.). Some universities and departments may require acknowledgement of other third-party proofreading help, so it is advisable to check with your module lecturers for specific expectations (University of Warwick, n.d.).
8. Mosaic plagiarism
Mosaic plagiarism, also named patchwork plagiarism, or patchwriting, involves combining phrases and text from multiple sources and integrating them into your own work without quotation marks or attribution (Turnitin, n.d.). In such cases, the plagiarised material is interwoven with the writer’s original content. For instance, this can occur when embedding a clause from a source into your sentence without acknowledgement (Kramer, 2022).
The advent of AI has given rise to a hybrid, more advanced form of patchwriting misconduct, also referred to as patchwriting 2.0 (Young, 2025). Students use GenAI as a complex paraphrasing or research tool, but fail to cite the sources properly. For example, a student might ask an AI to explain a concept using information from several articles, then directly copy the AI-generated answer into their paper without attributing ideas to the sources (Young, 2025).
9. Source-based plagiarism
Providing incorrect or incomplete information about sources makes them challenging or impossible to locate (Turnitin, n.d.). This type of plagiarism also includes situations in which a secondary source is referenced, but only the primary source from which the secondary source is derived is credited (Kramer, 2022). Other examples include citing incorrect sources or even fabricating sources altogether (Kramer, 2022). With the rise of generative AI tools that can create fictitious references or non-existent sources, an advanced form of source-based plagiarism has now emerged (Young, 2025).
10. Inaccurate citation
Failure to cite correctly–whether in quoted passages, in-text citations, footnotes, reference lists or bibliographies–can breach academic standards (University of Oxford, n.d.). It is essential to adhere to the referencing conventions specific to your discipline and to include only those sources that you have genuinely consulted (University of Oxford, n.d.).
11. Data plagiarism
Data plagiarism involves the falsification (altering existing data) or fabrication (inventing false data) of data or the inappropriate appropriation of someone else’s work (taking something for your own use without permission) (Turnitin, n.d.). These behaviours seriously undermine the credibility of the original research and can jeopardise the reputation of the researchers, institutions, or publishers involved (Turnitin, n.d.).
12. AI misuse or AI plagiarism
Using generative AI, AI-powered paraphrasing or rewriting tools to alter text or complete work without proper attribution is considered an academic offence (Turnitin, n.d.). AI-generated or AI-paraphrased content submitted as your own work is usually seen as a serious breach of integrity (University of Warwick, n.d.). Suppose you are permitted to use AI tools in your work but fail to disclose their use, provide a misleading description of them, or exceed the allowed amount outlined in your assessment information (Students’ Union UCL, 2025; University of Warwick, n.d.). In that case, it may lead to an academic misconduct investigation later.
13. Manual text modification
Manually altering text to mislead, deceive or evade plagiarism detection software, such as Turnitin, raises ethical concerns (Turnitin, n.d.). The intent behind these alterations is key. If modifications are made to improve the quality of one’s work, they are not only acceptable but commendable. However, it is misconduct to make modifications to evade detection tools.
14. Automated text modification
Unlike manual text modification performed by humans, automated text modification relies on software tools, such as text spinners and translation engines, to alter content originally written by someone else or generated by a Large Language Model (LLM) with the sole purpose of bypassing plagiarism detection (Turnitin, n.d.). Using automated software to replace words with synonyms, rephrase sentences, reword content, or translate text into another language and back to English in order to change phrasing is considered a form of academic dishonesty.
A more advanced form of automated text modification, AI paraphrasing, has recently emerged. Unlike basic text spinners that rely on simple algorithms, AI paraphrasing tools utilise cutting-edge algorithms to rephrase text whilst preserving its original meaning, resulting in content that appears more original (Young, 2025).
In addition to AI paraphrasing, a more deceptive and concerning trend that has arisen in 2025 is the use of AI bypassers or AI humanisers (Young, 2025), which are promoted to disguise AI-generated or AI-paraphrased content to evade AI detection software, ultimately concealing AI use from educators (Chechitelli, 2025). The automated modification of AI-generated or AI-paraphrased writing to make it appear human-written, with the intention of deceiving and actively hiding the AI use, is also unethical and may result in disciplinary action.
15. Self-plagiarism
Self-plagiarism, also called auto-plagiarism, duplicate plagiarism, or recycled plagiarism, refers to the reuse of your own previously submitted or published work, whether in part or in whole, for a new assignment, either for your current course, a different course, or another university, and is not permitted (University of Birmingham, 2024; University of Oxford, n.d.). Moreover, concurrently submitting the same work for multiple courses or institutions is regarded as auto-plagiarism (University of Oxford, n.d.).
Nevertheless, suppose your course rules explicitly allow you to reuse your previous work (University of Oxford, n.d.), or you are writing an essay that builds upon your previous work (University of Cambridge, n.d.-a). In that case, you must reference the original work as if it were authored by someone other than yourself.
16. Collusion
Engaging in unauthorised collaboration, failing to credit the assistance you received, or not following specific rules for group projects (University of Oxford, n.d.). This includes working with other students on an assignment that you then submit as your own individual work (Turnitin, n.d.). Both the student who provided assistance and the student who submitted the work may be held equally responsible for misconduct (University of Warwick, n.d.).
17. Contract cheating or ghostwriting
Engaging with a third party to complete assessed work and then claiming that work as your own, whether for pay, for free, or in exchange for something else (Turnitin, n.d.). Neither should you hire professional agencies to produce your work nor commission someone else to write it for you, even if that person agrees to do so (University of Oxford, n.d.). Submitting work that you downloaded from electronic sources, including essay repositories, either in whole or in part, also violates academic integrity policies (Imperial College London, 2023; University of Birmingham, 2024).
Furthermore, using GenAI (including Grammarly Premium’s rewriting and tone adjustment features) to produce phrases and sentences constitutes contract cheating. This is because it involves using AI to create writing or content that you then submit as your own work for assessment (University of Bristol, n.d.).
18. Unacceptable proofreading
Proofreading is generally permitted in UK universities and many other countries, whether carried out by professional or non-professional human proofreaders, or through AI tools, provided it adheres to academic integrity standards. Educational institutions typically provide clear guidelines outlining acceptable and unacceptable proofreading practices, allowing students to enhance their work while upholding ethical standards.
Acceptable proofreading includes corrections to grammar, spelling, punctuation, formatting and layout, and minor referencing errors. Proofreaders may also highlight unclear or complex sentences, repeated phrases or missing words (University of Nottingham, 2025b). Nonetheless, proofreaders may not add, edit, rewrite, rearrange, restructure content; change the content or meaning of the work, conduct fact-checking or data checking or correction, translate any work into English, make significant changes to the referencing system, and/or edit content to comply with world limits (University of Glasgow, 2024; University of Nottingham, 2025b).
Students are responsible for ensuring that any proofreading assistance they receive complies with academic integrity standards, whether it comes from a person, generative AI, or any other means. The use of GenAI for proofreading is generally discouraged because it can easily introduce mistakes, reinforcing the importance of human proofreaders (University of Glasgow, 2024).
AI proofreading or grammar tools are generally acceptable for basic spelling and grammar checks. For example, many students and academic staff use Grammarly to identify and correct these types of mistakes. Using AI tools for this purpose is seen as ethical, unless your assessment specifically assesses your ability to use correct grammar and spelling (University of Bristol, n.d.).
19. Falsification, fabrication or misrepresentation
A deliberate attempt to deceive the marker by presenting invented, false, negligent, misleading data, evidence, results, references, citations, AI-generated content (when explicitly prohibited by the assessment brief), or other material (University of Birmingham, 2024; University of Edinburgh, 2024; University of Nottingham, 2025a).
Such misconduct encompasses falsely claiming to have completed work, experiments, observations, interviews, or other forms of research that the student never actually undertook, as well as falsely claiming to have obtained results or other evidence (University of Nottingham, 2025a).
In addition, it involves falsification, fabrication, or misrepresentation of other research aspects, such as documentation and participant consent, presenting or recording such data, practice hours, required competencies, claims of extenuating circumstances, word counts, or assessement submission receipts, presenting them as if they were legitimate (UCL, n.d.-a; University of Cambridge, n.d.-b; University of Nottingham, 2025a).
20. Any other behaviours or actions that would give you an unfair academic advantage
Other dishonest acts that provide you with an unfair academic advantage are not explicitly mentioned above (Students’ Union UCL, 2025). For instance, offering a bribe or inducement to a member of academic or professional support staff (Imperial College London, 2023). In addition, when an assessment requires you to submit work in English. Still, you write in a different language and then use translation software to convert it into English to gain an unfair academic advantage (Heriot-Watt University, n.d.), among other dishonest practices.
Conclusion
As AI-driven misconduct becomes increasingly prevalent and sophisticated, staying informed about both traditional and emerging forms of plagiarism and academic misconduct is vital. This understanding can reduce opportunities for academic dishonesty, address these issues more effectively, and cultivate a strong commitment to ethical standards. To support these standards, rapid advancements in AI-powered plagiarism detection technologies, such as Turnitin, enable institutions and academic publishers to identify AI-generated, AI-paraphrased, and AI-humanised content more accurately, thereby ensuring accountability within the academic community (Chechitelli, 2025).
Nonetheless, technological detection is only one part of the solution. It is important to note that academic misconduct, plagiarism, and AI use policies can vary between universities, departments, and even among individual lecturers and supervisors. Therefore, students should always consult their university’s official academic integrity and AI use policies and seek clarification from their course lecturers and supervisors to fully understand expectations. This matters because similarity and AI writing reports are intended to assist academic judgement, not to serve as automatic verdicts; the final decision rests with the course lecturers and supervisors and the Academic Integrity Officers (also known as Academic Misconduct Officers).
Click to view a reference list of 20 types of plagiarism and academic misconduct.