Search for: All records

Recently, there have been significant advances and wide-scale use of generative AI in natural language generation. Models such as OpenAI’s GPT3 and Meta’s LLaMA are widely used in chatbots, to summarize documents, and to generate creative content. These advances raise concerns about abuses of these models, especially in social media settings, such as large-scale generation of disinformation, manipulation campaigns that use AI-generated content, and personalized scams. We used stylometry (the analysis of style in natural language text) to analyze the style of AI-generated text. Specifically, we applied an existing authorship verification (AV) model that can predict if two documents are written by the same author on texts generated by GPT2, GPT3, ChatGPT and LLaMA. Our AV model was trained only on human-written text and was effectively used in social media settings to analyze cases of abuse. We generated texts by providing the language models with fanfiction snippets and prompting them to complete the rest of it in the same writing style as the original snippet. We then applied the AV model across the texts generated by the language models and the human written texts to analyze the similarity of the writing styles between these texts. We found that texts generated with GPT2 had the highest similarity to the human texts. Texts generated by GPT3 and ChatGPT were very different from the human snippet, and were similar to each other. LLaMA-generated texts had some similarity to the original snippet but also has similarities with other LLaMA-generated texts and texts from other models. We then conducted a feature analysis to identify the features that drive these similarity scores. This analysis helped us answer questions like which features distinguish the language style of language models and humans, which features are different across different models, and how these linguistic features change over different language model versions. The dataset and the source code used in this analysis have been made public to allow for further analysis of new language models. 

Cryptographic tools for authenticating the provenance of web-based information are a promising approach to increasing trust in online news and information. However, making these tools’ technical assurances sufficiently usable for news consumers is essential to realizing their potential. We conduct an online study with 160 participants to investigate how the presentation (visual vs. textual) and location (on a news article page or a third-party site) of the provenance information affects news consumers’ perception of the content’s credibility and trustworthiness, as well as the usability of the tool itself. We find that although the visual presentation of provenance information is more challenging to adopt than its text-based counterpart, this approach leads its users to put more faith in the credibility and trustworthiness of digital news, especially when situated internally to the news article.