Detecting AI-Generated Text: Factors Influencing Detectability with Current Methods

2024-06-21

Journal of Artificial Intelligence Research

AI Summary

This work addresses the critical AI governance challenge of detecting AI-generated text (AIGT). We systematically survey state-of-the-art detection approaches—including watermarking, statistical feature analysis, stylistic modeling, and supervised/zero-shot machine learning classifiers—and catalog major benchmark datasets. We introduce, for the first time, a multidimensional framework characterizing AIGT detectability, incorporating factors such as LLM architecture, text length, and domain alignment. Further, we propose a practical, deployment-oriented detectability evaluation framework accompanied by actionable guidelines. Empirical analysis reveals fundamental performance limits and robustness bottlenecks of existing methods under cross-model and cross-domain settings. Our findings provide theoretical foundations and an implementable technical roadmap for applications in fraud detection, academic integrity assurance, and disinformation mitigation.

Technology Category

• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)
• Machine Learning: Large Multimodal Models (LMMs)
• Philosophy and Ethics of AI: Artificial General Intelligence

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Large language models (LLMs) have advanced to a point that even humans have difficulty discerning whether a text was generated by another human, or by a computer. However, knowing whether a text was produced by human or artificial intelligence (AI) is important to determining its trustworthiness, and has applications in many domains including detecting fraud and academic dishonesty, as well as combating the spread of misinformation and political propaganda. The task of AI-generated text (AIGT) detection is therefore both very challenging, and highly critical. In this survey, we summarize stateof-the art approaches to AIGT detection, including watermarking, statistical and stylistic analysis, and machine learning classification. We also provide information about existing datasets for this task. Synthesizing the research findings, we aim to provide insight into the salient factors that combine to determine how “detectable” AIGT text is under different scenarios, and to make practical recommendations for future work towards this significant technical and societal challenge.

Problem

• Detecting AI-generated text using current methods
• Assessing trustworthiness of human vs AI-written content
• Identifying factors influencing detectability in different scenarios

Innovation

• Watermarking for AI text detection
• Statistical and stylistic analysis methods
• Machine learning classification techniques