The Widespread Adoption of Large Language Model-Assisted Writing Across Society

2025/02/13 by Weixin Liang, Yaohui Zhang, Liang, Weixin +9 · 17 voices · 24 citations
Computer Science · #Topic Modeling #cs.CL

paper · pdf · doi:10.48550/arxiv.2502.09747

openalex publication_date 2025/02/13 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

The recent advances in large language models (LLMs) attracted significant public and policymaker interest in its adoption patterns. In this paper, we systematically analyze LLM-assisted writing across four domains-consumer complaints, corporate communications, job postings, and international organization press releases-from January 2022 to September 2024. Our dataset includes 687,241 consumer complaints, 537,413 corporate press releases, 304.3 million job postings, and 15,919 United Nations (UN) press releases. Using a robust population-level statistical framework, we find that LLM usage surged following the release of ChatGPT in November 2022. By late 2024, roughly 18% of financial consumer complaint text appears to be LLM-assisted, with adoption patterns spread broadly across regions and slightly higher in urban areas. For corporate press releases, up to 24% of the text is attributable to LLMs. In job postings, LLM-assisted writing accounts for just below 10% in small firms, and is even more common among younger firms. UN press releases also reflect this trend, with nearly 14% of content being generated or modified by LLMs. Although adoption climbed rapidly post-ChatGPT, growth appears to have stabilized by 2024, reflecting either saturation in LLM adoption or increasing subtlety of more advanced models. Our study shows the emergence of a new reality in which firms, consumers and even international organizations substantially rely on generative AI for communications.

Citations

• Monitoring AI-Modified Content at Scale: A Case Study on the Impact of ChatGPT on AI Conference Peer Reviews
• GPT detectors are biased against non-native English writers
• On the Opportunities and Risks of Foundation Models
• On the Dangers of Stochastic Parrots

Cited by

• Can Good Writing Be Generative? Expert-Level AI Writing Emerges through Fine-Tuning on High-Quality Books
• AI use in American newspapers is widespread, uneven, and rarely disclosed
• BitFlipScope: Scalable Fault Localization and Recovery for Bit-Flip Corruptions in LLMs
• HATS: High-Accuracy Triple-Set Watermarking for Large Language Models
• The American Ghost in the Machine: How language models align culturally and the effects of cultural prompting
• Training-Free Policy Violation Detection via Activation-Space Whitening in LLMs
• Large Language Models' Complicit Responses to Illicit Instructions across Socio-Legal Contexts
• AI-Assisted Writing Is Growing Fastest Among Non-English-Speaking and Less Established Scientists
• Cmprsr: Abstractive Token-Level Question-Agnostic Prompt Compressor
• Finding Diamonds in Conversation Haystacks: A Benchmark for Conversational Data Retrieval
• The Rise of Large Language Models and the Direction and Impact of US Federal Research Funding
• Future of AI Models: A Computational perspective on Model collapse
• In Generative AI We (Dis)Trust? Computational Analysis of Trust and Distrust in Reddit Discussions
• On the Detectability of LLM-Generated Text: What Exactly Is LLM-Generated Text?
• Measuring Moral LLM Responses in Multilingual Capacities
• Fine-Grained Uncertainty Decomposition in Large Language Models: A Spectral Approach
• Belief Coevolution in a Social Network of Generalist and Specialist Large Language Models
• Sympathetic Framing: Evaluating AI Alignment across Sociodemographic Groups
• Designing Culturally Aligned AI Systems For Social Good in Non-Western Contexts
• Can GenAI Move from Individual Use to Collaborative Work? Experiences, Challenges, and Opportunities of Integrating GenAI into Collaborative Newsroom Routines
• Vibe Coding for UX Design: Understanding UX Professionals' Perceptions of AI-Assisted Design and Development
• Political Ideology Shifts in Large Language Models
• LongRecall: A Structured Approach for Robust Recall Evaluation in Long-Form Text
• How Persuasive Could LLMs Be? A First Study Combining Linguistic-Rhetorical Analysis and User Experiments

Discussions

• Paper: arxiv.org/abs/2502.09747 [bsky, 28 points, 1 comments]
• Preprint from Liang et al showing about a quarter of corporate press releases--and smaller chunks of other genres--are now written with LLMs. This really is a question of genre, because the press rele [bsky, 15 points, 0 comments]
• Large language model-assisted writing is already prevalent, including in almost 14% of the UN press releases: arxiv.org/abs/2502.09747 [bsky, 7 points, 0 comments]
• AN 'AI slop' study found substantial AI text in corporate press releases (24%), job posts, even UN communications. This isn't just noise; it's a quantifiable shift in the internet's voice. Meh. Link t [bsky, 5 points, 2 comments]
• By late 2024, roughly 18% of financial consumer complaint text, 24% of corporate press releases, & 10% of small business job postings appears to be AI LLM-assisted arxiv.org/abs/2502.09747 [bsky, 3 points, 0 comments]
• "Although adoption climbed rapidly post-ChatGPT, growth appears to have stabilized by 2024, reflecting either saturation in LLM adoption or increasing subtlety of more advanced models." arxiv.org/abs [bsky, 3 points, 1 comments]
• 🚨 Our new study reveals widespread LLM adoption across society: 📊By late 2024, LLMs assist in writing: - 18% of financial consumer complaints - 24% of corporate press releases - Up to 15% of job p [bsky, 3 points, 0 comments]
• The Widespread Adoption of Large Language Model-Assisted Writing Across Society [hn, 2 points, 0 comments]
• The way we write is changing across society due to #AI #LLMs arxiv.org/abs/2502.09747 [bsky, 2 points, 1 comments]
• AI is already doing our work. As of September 2024, 18% of financial consumer complaints, 24% of corporate press releases, and 15% of job postings showed signs of LLM writing. arxiv.org/pdf/2502.09747 [bsky, 2 points, 0 comments]
• The Widespread Adoption of Large Language Model-Assisted Writing Across Society [hn, 1 points, 0 comments]
• We’re in the midst of a language speciation event and it’s not clear to me where it’s evolving “By. September 2024, 18% of financial consumer complaints, 24% of press releases, 15% of job postings & [bsky, 1 points, 0 comments]
• On the use of LLM's, the evidence tells us: “Our findings reveal widespread adoption of large language models across diverse writing domains, ranging consumers, firms and international organizations [bsky, 1 points, 0 comments]
• The Widespread Adoption of LLM-Assisted Writing Across Society [hn, 1 points, 0 comments]
• Researchers surprised to find less-educated areas adopting AI writing tools faster via @cornell arxiv.org/abs/2502.09747 #EduSky #EduSkyAI #EduCoach #TLSky #EdTech #AIinEducation #aisky #ai #artif [bsky, 0 points, 0 comments]
• Customer complaints, press releases, job ads are all increasingly being written with LLMs arxiv.org/abs/2502.09747 [bsky, 0 points, 0 comments]
• Studies show AI-generated text in 24% of press releases, 15% of job postings & even UN statements. 🤖 Small businesses, students & professionals are all using AI—but are we relying on it too much? ⚖️ [bsky, 0 points, 0 comments]

Related

• Computational Approaches to Understanding Large Language Model Impact on Writing and Information Ecosystems
• The Adoption and Efficacy of Large Language Models: Evidence From Consumer Complaints in the Financial Industry
• Visualizing the Adoption of Large Language Models across Sociology Subfields
• Mapping the Increasing Use of LLMs in Scientific Papers
• The diffusion of large language models in published academic articles