Automatic Detection of Machine Generated Text: A Critical Survey

By Ganesh Jawahar, Muhammad Abdul-Mageed, Laks V. S. Lakshmanan. 2020

305 citations recorded by the source corpus.

Text generative models (TGMs) excel in producing text that matches the style of human language reasonably well. Such TGMs can be misused by adversaries, e.g., by automatically generating fake news and fake product reviews that can look authentic and fool humans. Detectors that can distinguish text generated by TGM from human written text play a vital role in mitigating such misuse of TGMs. Recently, there has been a flurry of works from both natural language processing (NLP) and machine learning (ML) communities to build accurate detectors for English. Despite the importance of this problem, there is currently no work that surveys this fast-growing literature and introduces newcomers to important research challenges. In this work, we fill this void by providing a critical survey and review of this literature to facilitate a comprehensive understanding of this problem. We conduct an in-depth error analysis of the state-of-the-art detector and discuss research directions to guide future work in this exciting area.

Subfields: Artificial Intelligence, Sociology and Political Science, Natural Language Processing, Evaluation & Benchmarks, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Sociology and Political Science

In Natural Language Processing

In Evaluation & Benchmarks

In Signal Processing

Citations in the corpus

Language Generation Models Can Cause Harm: So What Can We Do About It? An Actionable Survey (2022, background): “Another commonly used approach is to detect whether text has been machine-generated (Dugan et al., 2020; Gehrmann et al., 2019), empowering users themselves to understand the source of the text they are reading (Jawahar et al., 2020).”
The Gradient of Generative AI Release: Methods and Considerations (2023, methodology): “Approaches to detection can be tailored to modality [42] such as text [35], and include human annotation [29].”
Knowledge Conflicts for LLMs: A Survey (2024, unknown): “Stochastic sampling methods like top-k and top-p sampling are commonly used for decoding, introducing randomness in the generated content (Jawahar et al., 2020; Massarelli et al., 2020; Fan et al., 2018; Holtzman et al., 2020).”
A Watermark for Large Language Models (2023, unknown): “…to watermarking is to develop detection models that perform a post-hoc analysis of machine-generated text, for example using language model features or ﬁnetuning existing large language models to behave as detectors (Zellers et al., 2019; Tan et al., 2020), see an overview in Jawahar et al. (2020).”
mGPT: Few-Shot Learners Go Multilingual (2022, background): “, spreading propaganda and misinformation, abusing legal and governmental procedures, and facilitating fraud, scams, and other targeted manipulation (Jawahar et al., 2020).”
AI Transparency in the Age of LLMs: A Human-Centered Research Roadmap (2023, background): “There is also an active line of research on post-hoc detection of artificially generated text (Jawahar et al., 2020; Tan et al., 2020; Zellers et al., 2019).”
Fast-DetectGPT: Efficient Zero-Shot Detection of Machine-Generated Text via Conditional Probability Curvature (2023, background): “Current research primarily concentrates on supervised methods , involving the training of classifiers to distinguish between machine-generated and human-written text (Jawahar et al., 2020).”
Can AI-Generated Text be Reliably Detected? (2023, background): “Several detection works study this problem as a binary classiﬁcation problem [3, 14, 2, 15, 16].”
LLM Evaluators Recognize and Favor Their Own Generations (2024, unknown): “Detection of LLM-generated text is important to both AI safety and combating misinformation (Jawahar et al., 2020; Crothers et al., 2023; Wu et al., 2023; Yang et al., 2023;”
DetectLLM: Leveraging Log Rank Information for Zero-Shot Detection of Machine-Generated Text (2023, background): “The problem of distinguishing machine-generated from human-written text is commonly formulated as a binary task (Jawahar et al., 2020).”