In the era of generative AI like ChatGPT, Claude, and Gemini, academic misconduct is changing drastically: While the number of violations isn’t increasing significantly, their form is becoming more sophisticated. Where previously texts were purchased from paper mills or copied directly from the internet, content today is often created within seconds using AI tools. A simple “humanizer” or paraphrasing quickly makes these texts invisible to traditional plagiarism software.
But there is a reliable alternative: Writing process analysis.
The Problem: New Technologies, Old Temptations
Academic misconduct is not a new phenomenon. Even before the introduction of ChatGPT in November 2022, between 60% and 70% of students admitted to having cheated at least once. These numbers have not changed significantly since then.
What has changed dramatically, however, is the method of cheating. Instead of “googling” content, obtaining it from paper mills, or hiring ghostwriters, learners today increasingly turn to generative AI. These create seemingly authentic texts within the shortest time. A survey found that around 90% of college students have already used AI for their homework.
This development presents new challenges for educational institutions. Traditional plagiarism detection software is not built to identify AI-generated content. As a first step, providers have extended these with ‘AI detection’ designed to find patterns of AI-generated texts. “Humanizer” tools and simple paraphrasing prompts already make this content practically undetectable.
The result: AI detection software that focuses on end products proves unreliable – a fact that has been clearly demonstrated in several studies (for details, see: AI Detection Tools? When You Turn It In, It’s Too Late!).
To address these challenges, new approaches are needed that go beyond analyzing the end product and examine the entire writing process.
The Revolution: Writing Process Analysis by Mentafy
This is where Mentafy’s innovative solution comes in. Our writing process analysis uses digital forensics to determine whether text was copied into a document, typed, or actually “conceived by humans before the machine.”
The paradigm shift is simple but radical: Deploy software from the beginning of the writing project, rather than just checking the end product.
As Clare Johnson aptly states in her groundbreaking dissertation: “The writing process itself is as important (if not more important) than the end product, because it is during the writing stage that the student will learn the most.”
Why Does This Work?
The human writing process follows characteristic patterns regarding:
- Tempo and rhythm of writing
- Correction patterns and revision behavior
- Thinking pauses and natural interruptions
- Individual writing habits
These biometric fingerprints of writing cannot be simulated by any known TyperScript or AI tool – deceiving the analysis is therefore extremely complex.
Privacy by Design
Our system captures only the data necessary for analysis and deletes everything else during report generation. This ensures full GDPR compliance and follows the Goldilocks principle: As much as necessary, as little as possible.
The Added Value: Learning Instead of Just Evaluating
The true innovation lies not only in monitoring, but in the learning support that writing process analysis simultaneously enables:
- Immediate feedback: Currently, we offer real-time feedback on correct citation.
- Future outlook: From mid-2026, we plan to introduce a complete writing tutor for all who truly want to learn academic writing.
Why Written Work Remains Crucial
It is often suggested in this context to introduce new examination forms or modify existing ones. There is much to be said for preserving academic writing as a central component in the education of our future academics. Written examination forms such as essays, papers, and theses have crucial pedagogical advantages:
- They promote critical thinking and independent research.
- They strengthen problem-solving skills and deep understanding of academic content.
- They provide equal opportunities through transparent evaluation of individual performance.
Writing process analysis enables educational institutions to maintain these important examination forms fairly and effectively.
The Appeal: Abolish Plagiarism Checkers and Replace Them with Writing Process Analysis
The time is ripe for fundamental change. Universities and educational institutions should:
- Replace outdated plagiarism software with process-based tools
- Enable fair evaluation of actual individual performance
- Consistently maintain learning objectives
- Preserve written work as a valuable form of learning
Mentafy’s writing process analysis offers more than just fraud detection – it enables a renaissance of authentic academic writing. In a world where AI-generated content is ubiquitous, the ability to document and evaluate the human thought process becomes a crucial didactic competency.
The future of academic integrity lies not in the perfect detection of AI texts, but in understanding and appreciating the human writing process itself: Writing with AI but also with brains!
Learn more about Mentafy’s innovative writing process analysis and how it can help your educational institution redefine academic integrity.









