Open Access

ARTICLE

Upholding Academic Integrity amidst Advanced Language Models: Evaluating BiLSTM Networks with GloVe Embeddings for Detecting AI-Generated Scientific Abstracts

Lilia-Eliana Popescu-Apreutesei, Mihai-Sorin Iosupescu, Sabina Cristiana Necula, Vasile-Daniel Păvăloaia^*

Department of Accounting, Information Systems and Statistics, Faculty of Economics and Business Administration, Alexandru Ioan Cuza University of Iasi, Iasi, 700505, Romania

* Corresponding Author: Vasile-Daniel Păvăloaia. Email:

(This article belongs to the Special Issue: Enhancing AI Applications through NLP and LLM Integration)

Computers, Materials & Continua 2025, 84(2), 2605-2644. https://doi.org/10.32604/cmc.2025.064747

Received 22 February 2025; Accepted 12 May 2025; Issue published 03 July 2025

Abstract

The increasing fluency of advanced language models, such as GPT-3.5, GPT-4, and the recently introduced DeepSeek, challenges the ability to distinguish between human-authored and AI-generated academic writing. This situation is raising significant concerns regarding the integrity and authenticity of academic work. In light of the above, the current research evaluates the effectiveness of Bidirectional Long Short-Term Memory (BiLSTM) networks enhanced with pre-trained GloVe (Global Vectors for Word Representation) embeddings to detect AI-generated scientific abstracts drawn from the AI-GA (Artificial Intelligence Generated Abstracts) dataset. Two core BiLSTM variants were assessed: a single-layer approach and a dual-layer design, each tested under static or adaptive embeddings. The single-layer model achieved nearly 97% accuracy with trainable GloVe, occasionally surpassing the deeper model. Despite these gains, neither configuration fully matched the 98.7% benchmark set by an earlier LSTM-Word2Vec pipeline. Some runs were over-fitted when embeddings were fine-tuned, whereas static embeddings offered a slightly lower yet stable accuracy of around 96%. This lingering gap reinforces a key ethical and procedural concern: relying solely on automated tools, such as Turnitin’s AI-detection features, to penalize individuals’ risks and unjust outcomes. Misclassifications, whether legitimate work is misread as AI-generated or engineered text, evade detection, demonstrating that these classifiers should not stand as the sole arbiters of authenticity. A more comprehensive approach is warranted, one which weaves model outputs into a systematic process supported by expert judgment and institutional guidelines designed to protect originality.

---

Keywords

---