Writing Style of Claude
|
|
|
Introduction. Several articles in popular journals [1-4] have identified telltale signs of AI writing. However, these articles have focused on essay writing and journalism. What about technical writing? This webpage analyzes the style of report summaries generated with Claude: Sonnet 5 (Medium). Although the summaries were generated with AI, the technical reports that were summarized had been written by top students using traditional methods.
Methods. To provide a reference for the style of Claude-generated summaries, we used averaged characteristics of style from the student-written reports. These reports, which Claude summarized, documented the initial phase of a course's design project that spanned the semester. What distinguished this course was its coupling of an engineering design course with a technical writing course [5]. From more than 300 reports written over a five-year period, the ten reports used in the study were chosen because of their technical precision and writing quality [6]. Each report documented one of four different projects: a line following robot [7], an autonomous etch-a-sketch [8], a training device for a service dog [9], or an exhibit for a children's science museum [10]. We chose a report summary as our AI document to test because having an AI tool generate a summary from a trusted report, as opposed to allowing the AI tool to generate text from its own sources, theoretically reduces the likelihood of imprecision.** Research has shown that such imprecision in technical writing arises from hallucination [11], exaggeration [12], and reference listings that either do not exist or do not support the given claim [13]. In addition, having a length limit on the summaries (in our case, no more than 1000 characters) theoretically reduces the likelihood of workslop [14-15]. To determine the baseline style of Claude, our prompt for the summary did not provide any instructions about the style for the summary. Our prompt provided the context of the writing situation, the role of the AI tool in the document, the action ("write a summary"), and the length (no more than 1000 characters). Moreover, we used a private chat window to avoid any effects from prior chats with the AI tool. While Claude was free to adopt the style of the report being summarized, we did not instruct the AI tool to do so. Results. In our analysis of the style of the summaries generated by Claude, we found strengths and weaknesses. The main strength of the technical summaries generated by Claude (Sonnet 5) was that these summaries conveyed the most important details of the respective reports [16]. In addition, all the sentences in the summaries added value for readers. Put another way, the summaries did not contain noticeable workslop. Three other strengths were that the summaries had a logical order, used active-voice verbs about 75 percent of the time, and in general were grammatically correct. As shown in Tables 1 and 2, Claude had distinct weaknesses. First, as shown in the AI-generated summaries were more difficult to read than the original reports were. This increased difficulty for reading the ChatGPT-generated summaries revealed itself in the following statistics:
|
|
|
Other weaknesses included sometimes being imprecise, such as incorrectly stating the scope of the report. A grammatical error that sometimes occurred was misplacing a participial phrase that modified a noun.
Besides these weaknesses and strengths, we noticed that distinctive aspects of the Claude-generated text occurred that were neither a weakness nor a strength:
Conclusions. The Claude-generated summaries contained the most important details of the reports, but the style was relatively complex and was restricted essentially to three types of sentence openers. Our next step is to determine what prompts, if any, could address these weaknesses. |
References
- Editors of The Economist, "How to Spot AI Writing: Large language models like long words and em-dashes--or do they?" Culture: Paper Trails, The Economist (30 July 2026).
- Sam Kriss, "Why Does A.I. Write Like...That?" The New York Times Magazine (3 December 2025).
- Eve Fairbanks, "The Biggest Tell That Something Was Written by AI," The Atlantic (29 May 2026).
- Charlie Fink, "The Seven Deadly Tells of AI Writing," Forbes (25 June 2025).
- Michael Alley, Stephanie Cutler, J. C. Tice, "Work-in-Progress: Embedding a large writing course in engineering design—a new model to teach technical writing," 2019 ASEE National Conference (Tampa: ASEE, 16 June 2019), DOI 10.18260/1-2--33610.
- Michael Alley, The Craft of Scientific Writing, 4th ed. (New York: Springer, 2018).
- Samantha Avery, Margaret Cross, and Robert Ritter, "Initial Research for the Design of an Autonomously Navigating Robot," report (University Park, PA: ME 340, Pennsylvania State University, 3 April 2023).
- Christine Gabriele, Arianna diPalma, Andrew Banks, and Sandon Birch, "Initial Research for the Design of an Autonomous Etch-a-Sketch," report (University Park, PA: ME 340, Pennsylvania State University, 22 February 2021).
- Joe Keaton, Leo Randolph, Hady Diab, and Gavin Nika, "Initial Research for the Design of Wheelchair-Operated Food Dispenser for Canine Training," report (University Park, PA: ME 340, Pennsylvania State University, 5 February 2025).
- Andrea Khouzami, Maria Lazzaro, and Ryan Galella, "General Design Concepts for a Children's Science Exhibit at the Discovery Space Museum," report (University Park, PA: ME 340, Pennsylvania State University, 20 November 2025).
- Yiqiu Shen, Laura Heacock, Keith D. Hentel, Beautriu Reig, George Shih, and Linda Moy, "ChatGPT and Other Large Language Models Are Double-edged Swords," Radiology, vol. 37, no. 2 (26 January 2023).
- Sai Anirudh Athaluri, Sandeep Varma Manthema, V. S. R. Krishna Manoj Kesapranda, Vineel Yarlagadda, Tirth Dave, and Rama Tulasi Siri Duffumpudi, "Exploring the Boundaries of Reality: Investigating the Phenomenon of Artificial Intelligence Hallucination in Scientific Writing Through ChatGPT References," Cureus (11 April 2023), DOI: 10.7759/cureus.37432.
- Smriti Mallapaty, "Signs of AI-generated text found in 14% of biomedical abstracts last year," Nature (2 July 2025).
- BetterUp, "Workslop is the new busywork, and it's costing millions," betterup.com/workslop" (Austin, Texas: BetterUp, 2026).
- Kate Niederhoffer, Gabriella Rosen Kellerman, Angela Lee, Alex Aliebscher, Kristina Rapuano, and Jeffrey T. Hancock, "AI-Generated "Workshop" Is Destroying Productivity," Harvard Business Review (22 September 2025).
- Michael Alley, Hala Hobrom, and Suryansh Sijwali, “Three Stylistic Weaknesses of AI Writing in Engineering Documents and Prompts to Address Those Weaknesses,” extended abstract, 2026 ASEE St. Lawrence Regional Conference (Ithaca, NY: ASEE St. Lawrence Regional Conference, 25 April 2026).
* Other than the AI example in Figure 2, AI was not used to generate text for this webpage. However, we did use ChatGPT 5.3 to assist in calculations of statistical significance.
**Colors are used in on this page to reflect different types of stylistic and grammatical signatures: imprecision, active voice, subject sentence opener, participle phrase, and needless complexity.
**Colors are used in on this page to reflect different types of stylistic and grammatical signatures: imprecision, active voice, subject sentence opener, participle phrase, and needless complexity.
