Two Birds with One Stone: Multi-Task Detection and Attribution of LLM-Generated Text

TL;DR

This paper introduces DA-MTL, a multi-task learning framework that effectively detects AI-generated text and attributes it to specific LLMs across multiple languages and models, enhancing security and forensic analysis.

Contribution

The paper presents a novel multi-task learning approach that jointly performs detection and authorship attribution for LLM-generated text, improving performance and robustness over existing methods.

Findings

Strong performance across nine datasets and four models

Effective in multiple languages and LLM sources

Robust against adversarial obfuscation techniques

Abstract

Large Language Models (LLMs), such as GPT-4 and Llama, have demonstrated remarkable abilities in generating natural language. However, they also pose security and integrity challenges. Existing countermeasures primarily focus on distinguishing AI-generated content from human-written text, with most solutions tailored for English. Meanwhile, authorship attribution--determining which specific LLM produced a given text--has received comparatively little attention despite its importance in forensic analysis. In this paper, we present DA-MTL, a multi-task learning framework that simultaneously addresses both text detection and authorship attribution. We evaluate DA-MTL on nine datasets and four backbone models, demonstrating its strong performance across multiple languages and LLM sources. Our framework captures each task's unique characteristics and shares insights between them, which boosts performance in both tasks. Additionally, we conduct a thorough analysis of cross-modal and cross-lingual patterns and assess the framework's robustness against adversarial obfuscation techniques. Our findings offer valuable insights into LLM behavior and the generalization of both detection and authorship attribution.