Automatic Detection of Machine Generated Text: A Critical Survey

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

doi:10.48550/arxiv.2011.01314

Cited by 21 publications

(27 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Fake news generated by simpler language models were also hard to detect and found to pass as human (Zellers et al, 2020). The risk of fake news generated by LMs is widely recognised and has spurred research into detecting such synthetic content (Jawahar et al, 2020). On polarisation, (McGuffie and Newhouse, 2020) demonstrated that via simple prompt engineering, GPT-3 can be used to generate content that emulates content produced by violent far-right extremist communities.…”

Section: Examplesmentioning

confidence: 99%

Ethical and social risks of harm from Language Models

Weidinger¹,

Mellor²,

Rauh³

et al. 2021

Preprint

105

View full text Add to dashboard Cite

This paper aims to help structure the risk landscape associated with large-scale Language Models (LMs). In order to foster advances in responsible innovation, an in-depth understanding of the potential risks posed by these models is needed. A wide range of established and anticipated risks are analysed in detail, drawing on multidisciplinary literature from computer science, linguistics, and social sciences.

show abstract

Section: Examplesmentioning

confidence: 99%

Ethical and social risks of harm from Language Models

Weidinger¹,

Mellor²,

Rauh³

et al. 2021

Preprint

105

View full text Add to dashboard Cite

show abstract

“…Moreover, (Uchendu et al, 2020) shows that RoBERTa outperforms existing detectors in detecting automatically generated news articles and product reviews which are generated by state of the art models like GPT-2. Despite the success of RoBERTa, recent research (Jawahar et al, 2020) shows that its dependence on large amounts of data, limits limits its use for detection. (Wolff and Wolff, 2020) challenges the RoBERTa model by exposing it with homoglyph and misspelling attacks and their results show a drastic drop in recall.…”

Section: Automatic Detection Of Machine Generated Textmentioning

confidence: 99%

A Benchmark Corpus for the Detection of Automatically Generated Text in Academic Publications

Liyanage¹,

Buscaldi²,

Nazarenko³

2022

Preprint

View full text Add to dashboard Cite

Automatic text generation based on neural language models has achieved performance levels that make the generated text almost indistinguishable from those written by humans. Despite the value that text generation can have in various applications, it can also be employed for malicious tasks. The diffusion of such practices represent a threat to the quality of academic publishing. To address these problems, we propose in this paper two datasets comprised of artificially generated research content: a completely synthetic dataset and a partial text substitution dataset. In the first case, the content is completely generated by the GPT-2 model after a short prompt extracted from original papers. The partial or hybrid dataset is created by replacing several sentences of abstracts with sentences that are generated by the Arxiv-NLP model. We evaluate the quality of the datasets comparing the generated texts to aligned original texts using fluency metrics such as BLEU and ROUGE. The more natural the artificial texts seem, the more difficult they are to detect and the better is the benchmark. We also evaluate the difficulty of the task of distinguishing original from generated text by using state-of-the-art classification models.

show abstract

“…Whilst it is true that existing work has been conducted to develop models capable of detecting AI-generated texts, it is worth noting that these solutions are typically limited in their generalisability. Generally, successful models are only effective in detecting AI-generated texts from a specific known model (e.g., GPT-2, XLM) of a specific type (e.g., news articles, blog posts) (Jawahar, Abdul-Mageed, and Lakshmanan 2020). There exists no "silver bullet" capable of making these potential deceptive texts trivial to identify.…”

Section: Our Contributionsmentioning

confidence: 99%

“…However, it is important to note that currently it is still possible to distinguish between AI-generated texts and human-crafted texts when one is actively looking for them, with classifiers trained to this task achieving fair-to-good performances (Jawahar, Abdul-Mageed, and Lakshmanan 2020). Given this, it is likely that AA systems in current use can mitigate the threats of NLG-based deception by being combined with some form of classifier trained specifically to the task of detecting AI-generated content.…”

Section: Broader Perspectives and Ethicsmentioning

confidence: 99%

Are You Robert or RoBERTa? Deceiving Online Authorship Attribution Models Using Neural Text Generators

Jones¹,

Nurse²,

Li³

2022

Preprint

View full text Add to dashboard Cite

Recently, there has been a rise in the development of powerful pre-trained natural language models, including GPT-2, Grover, and XLM. These models have shown state-of-theart capabilities towards a variety of different NLP tasks, including question answering, content summarisation, and text generation. Alongside this, there have been many studies focused on online authorship attribution (AA). That is, the use of models to identify the authors of online texts. Given the power of natural language models in generating convincing texts, this paper examines the degree to which these language models can generate texts capable of deceiving online AA models. Experimenting with both blog and Twitter data, we utilise GPT-2 language models to generate texts using the existing posts of online users. We then examine whether these AI-based text generators are capable of mimicking authorial style to such a degree that they can deceive typical AA models. From this, we find that current AI-based text generators are able to successfully mimic authorship, showing capabilities towards this on both datasets. Our findings, in turn, highlight the current capacity of powerful natural language models to generate original online posts capable of mimicking authorial style sufficiently to deceive popular AA methods; a key finding given the proposed role of AA in real world applications such as spam-detection and forensic investigation.

show abstract

Automatic Detection of Machine Generated Text: A Critical Survey

Cited by 21 publications

References 0 publications

Ethical and social risks of harm from Language Models

Ethical and social risks of harm from Language Models

A Benchmark Corpus for the Detection of Automatically Generated Text in Academic Publications

Are You Robert or RoBERTa? Deceiving Online Authorship Attribution Models Using Neural Text Generators

Contact Info

Product

Resources

About