Can ChatGPT Detect DeepFakes? A Study of Using Multimodal Large Language Models for Media Forensics

TL;DR

This paper explores the potential of multimodal large language models to detect AI-generated DeepFake media, demonstrating their capabilities through experiments and discussing their limitations and possible enhancements.

Contribution

It is the first to evaluate multimodal LLMs for media forensics, showing they can identify DeepFakes without specialized programming.

Findings

Multimodal LLMs can detect DeepFake images with careful prompt design.

The models reveal AI-generated media through qualitative and quantitative tests.

Limitations of current models are identified and discussed.

Abstract

DeepFakes, which refer to AI-generated media content, have become an increasing concern due to their use as a means for disinformation. Detecting DeepFakes is currently solved with programmed machine learning algorithms. In this work, we investigate the capabilities of multimodal large language models (LLMs) in DeepFake detection. We conducted qualitative and quantitative experiments to demonstrate multimodal LLMs and show that they can expose AI-generated images through careful experimental design and prompt engineering. This is interesting, considering that LLMs are not inherently tailored for media forensic tasks, and the process does not require programming. We discuss the limitations of multimodal LLMs for these tasks and suggest possible improvements.