Earticle

심층위조는 새로운 인공지능 기술로 현재 운용 중에 있어 법적 위험과 새로운 법률 문제가 나타나고 있다. 이는 비전통적인 법률형태로 국가와 인민대중에게 안보 위험을 초래하고 있다. 현재 각국의 심층위조기술에 대한 형법 제정은 인공지능 정책과 기술 발전의 응용 상황에 따라 차이가 있다. 중국의 “14.5” 규획에 입각하여 국가안보를 보호 하는 전략목표를 유지하고 있다. 본 연구에서는 심층위조와 관련한 보안이론을 검토한다. 특히, 국내외 각국의 심층 위조기술에 대한 형법규제를 비교한다. 이에는 심층위조기술의 대상, 응용과정 및 사 용주체의 차원에 입각하여 중국의 심층위조기술의 형법규제 현황에 대한 개인생체식 별정보의 형법보호 강화, 심층위조기술 관련 죄목규제 증설 및 인터넷 플랫폼의 형사 책임에 관한 대책을 제시하였다. 이를 통해 형법규제를 통한 심층위조기술의 부정적인 영향을 억제하고, 긍정적 효과를 발휘하는데 좋은 자료로 활용될 것이라 본다.

Deep forgery technology is a new AI technology, which has brought security risks to the country and the people in non-traditional forms in the current application, and has emerged a variety of legal risks and legal issues that deserve attention. It has become the common trend of criminal law in various countries to use the criminal law anti system depth forgery technology to bring negative effects on public security and national security. Therefore, it is of great practical significance and theoretical value to summarize the international experience of criminal law in regulating deep forgery technology. Due to different national conditions and technology development and application, there are differences in criminal law regulation paths for deep forgery technology in various countries. Based on the strategic goal of maintaining and shaping national security in China’s “Fourteenth Five Year Plan”, this paper uses the security theory to compare the criminal law regulation paths of countries at home and abroad for deep forgery technology. From the three levels of the object, application process and user of the deep forgery technology, this paper compares the experience of various countries in applying criminal law to the deep forgery technology, and obtains the common characteristics of countries in criminal law regulation of the deep forgery technology. Through comparative analysis, this paper proposes three countermeasures to improve the criminal law regulation of deep forgery technology: first, strengthen the criminal law protection of personal biometric information, second, add the regulation of charges related to deep forgery technology, and third, clarify the criminal responsibility of the network platform. Through the improvement of the criminal law regulation path, it is beneficial to curb the negative impact of deep forgery technology, and maintain the bottom line of public security and national security.

ﾠ深度伪造技术是一种新兴的人工智能技术，在目前的运用中已经以非传统的形式给 国家和人民群众带来安全风险，已经出现了种种值得关注的法律风险和法律问题。动用 刑法反制深度伪造技术给公共安全和国家安全带来的负面影响，已经成为各国刑法的共 同趋势。因此，对刑法规制深度伪造技术的国际经验予以总结，具有重要的现实意义和 理论价值。 由于国情以及技术发展应用情况不同，各国目前针对深度伪造技术的刑法规制路径 存在差异。立足于中国“十四五”规划之中维护和塑造国家安全的战略目标，运用安全化 理论对海内外各国针对深度伪造技术刑法规制路径进行比较。从深度伪造技术的对象、 应用过程以及使用主体三个层面，对各国应用刑法对深度伪造技术的经验进行了比较， 得出了各国针对深度伪造技术予以刑法规制的共同特征。通过比较分析，本文提出了完 善深度伪造技术刑法规制的三点对策：一是强化个人生物识别信息的刑法保护，二是增 设深度伪造技术相关罪名规制，三是明确网络平台的刑事责任。通过刑法规制路径的完 善，有利于抑制深度伪造技术的消极影响，守住维护公共安全和国家安全的底线。

Audio recordings are used widely as crucial evidence due to the popularity of smartphones. However, the digital evidence submitted voluntarily may be questioned for its authentication. Since most smartphones provide the function of digital editing files, they could be exposed to crime as being edited maliciously. Primarily, it could be severe in the case of voluntary submission via information and communications networks. This paper proposed the forensic method for authentication of audio recordings generated by the Voice Recorder application from Samsung smartphones. First, we analyzed and compared the file structures and timestamps of audio recordings between the normal state without manipulation and the manipulated state. Then, we analyzed the spectrogram of audio recordings. Finally, by analyzing the timestamps of the media-log for audio recordings, we confirmed the authentication method for the audio recordings forensically.

We examine the weaknesses of the existing OOXML-based MS-Word file structure, and analyze how data concealment and forgery are performed in MS-Word digital documents. In case of forgery by including hidden information in MS-Word digital document, there is no difference in opening the file with the MS-Word Processor. However, the computer system may be malfunctioned by malware or shell code hidden in the digital document. If a malicious image file or ZIP file is hidden in the document by using the structural vulnerability of the MS-Word document, it may be infected by ransomware that encrypts the entire file on the disk even if the MS-Word file is normally executed. Therefore, it is necessary to analyze forgery and alteration of digital document through internal structure analysis of MS-Word file. In this paper, we designed and implemented a mechanism to detect this efficiently and automatic detection software, and presented a method to proactively respond to attacks such as ransomware exploiting MS-Word security vulnerabilities.

In the era of today's world verification of images plays an important role. Various image editing tools are available in the market which can make changes in image in different ways. By using these tools, we can make slight changes in the image by resizing, rotating, noise addition and stretching an image or by splicing or copy move which is difficult to detect by human eyes. The digital images plays important role in many fields such as criminal and forensic investigation, military, journalism etc. So we needed some forgery detection technique for digital image. This paper reviews techniques for pixel-based forgery detection. First is copy-move or cloning and second is fast-copy move detection. In copy-move or cloning technique a part of the image is copied and pasted into another part of the image. But this image has limitation of only shifting of copied regions. So, second technique fast-copy move detection is discussed. Its result is complex but precise. Main disadvantage of fast-copy move technique it is not being able to detect for very small region.

오늘날 누구나 디지털 정보를 용이하게 활용할 수 있게 됨에 따라 디지털 증거의 채택이 증가되고 있다. 하지 만 다양한 음성 파일 편집 도구를 보급과 함께 정교한 편집 과정을 거친 음성 녹음 파일의 경우 위변조 진위 여부를 판 단하는 것은 사실상 불가능하다. 본 연구는 음성 녹음 파일에 삽입, 삭제, 연결 및 합성 편집 기술을 활용해 원본 파 일과 구별하기 어려운 위변조가 가능함을 증명하고자 한다. 본 연구는 위변조 된 음성 파일을 원본과 동일한 확장자 로 인코딩하는 작업을 통해 위변조 검출의 어려움을 제시한다. 또한 특징점이 발생한 실험에 한 하여 추가적으로 천 이대역의 삭제 및 2차 인코딩 작업을 수행할 경우 위변조 검출은 불가능함을 나타냈다. 이를 통해 본 연구는 음성 녹 음 파일을 디지털 증거로 채택하기 위한 더 엄격한 증거능력 판단 기준 수립에 공헌할 것으로 기대된다.

As digital information is readily available to everyone today, the adoption of digital evidence is increasing. However, it is virtually impossible to determine the authenticity of forgery in the case of a voice recording file that has gone through a sophisticated editing process along with the spread of various voice file editing tools. This study aims to prove that forgery, which is difficult to distinguish from the original file, is possible by using insertion, deletion, linking, and synthetic editing technologies in voice recording files. This study presents the difficulty of detecting forgery by encoding a forged voice file with the same extension as the original. In addition, it was shown that forgery detection is impossible if additional transition band deletion and secondary encoding are performed only for experiments in which features occurred. Through this, this study is expected to contribute to the establishment of more stringent evidence admissibility criteria for adopting voice recording files as digital evidence.

소프트웨어 공급망 보안의 중요성이 증가함에 따라, 소프트웨어 자재 명세서(Software Bill of Materials, SBOM)의 신뢰성과 일관성을 확보하는 것이 핵심 과제로 대두되고 있다. 본 연구는 동일 프로젝트의 다양한 버전에서 생성된 SBOM 문서 간 유사도 분석을 통해 위조 가능성을 탐지하고, 변경 이력을 자동으로 분석하는 기법을 제안한다. 이를 위해 공개 SBOM 데이터셋에 대해, 사전 학습된 언어 모델을 이용하여 각 문서를 임베딩한 후 코사인 유사도 기반으로 비교하였다. 이후 OpenAI GPT-4o를 활용한 거대언어모델 기반 에이전트를 통해 실제 변경된 구성 요소를 식별하고, 사용자가 이해할 수 있는 설명 보고서를 생성하였다. 실험 결과, 본 기법은 기존의 단순 비교 방식에 비해 위조 탐지 정확도와 변경 이력 해석 측면에서 우수한 성능을 보였다. 본 연구는 SBOM 기반 보안 검사를 자동화하고, 신뢰 가능한 변경 이력 관리 도구 개발의 초석을 마련하는 데 기여할 수 있을 것으로 기대된다.

As the importance of software supply chain security continues to grow, ensuring the reliability and consistency of Software Bills of Materials (SBOM) has emerged as a critical challenge. In this study, we propose an automated method to detect potential forgeries and analyze component-level modifications by analyzing the semantic similarity between SBOM documents generated from different versions of the same software project. Using a publicly available SBOM dataset, each document is embedded with a pre-trained language model and compared using cosine similarity. Subsequently, structural differences are identified through dedicated large language model (LLM)-based agents, implemented via OpenAI GPT-4o, which generate human-readable explanation reports that highlight suspicious changes. Experimental results demonstrate that the proposed method outperforms traditional comparison techniques in both forgery detection accuracy and interpretability of change history. This research contributes to the automation of SBOM-based security auditing and provides a foundational approach for the development of trustworthy change tracking tools in the software supply chain domain.

오늘날 음성 파일은 전화 통화, 음성 메시지, 음성 녹음 등 다양한 방법으로 손쉽게 생성되면서 법정에서 증거물로 제출되는 경우가 점차적으로 증가하고 있다. 그러나 디지털 파일의 특성상 위변조가 가능해 악의적 목적을 가진 개인이나 조직에 의해 악용될 수 있는 위험이 존재한다. 따라서 음성 파일의 무결성과 신뢰성을 보장하기 위한 위변조 검출 기법의 중요성이 증대되고 있다. 본 논문은 음성 파일 위변조 검출 기법에 대한 국내 연구의 분석을 통해 한계점을 조사하고 발전 방향을 제안하는 것을 목적으로 한다. 음성 파일 위변조 검출 기법은 크게 주파수 분석, 메타데이터 및 파일 구조 분석, 인공지능 활용으로 분류된다. 이러한 세 가지 검출 기법들은 정밀한 편집의 경우에 음성 파일 위변조를 검출하는 데 한계를 지닌다. 따라서 본 연구는 음성 파일의 위변조 검출을 위한 딥러닝 모델 구축을 위한 새로운 알고리즘 개발 및 데이터 셋 구축의 필요성을 제안한다. 또한 본 연구는 위변조된 음성 파일이 법원에 증거물로 제출되기 전, 무결성을 입증하기 위한 딥러닝 기반의 인증 시스템의 필요성을 제안한다. 본 연구는 음성 파일 위변조 검출 기법에 대한 한계점 분석 및 발전 방향을 제시함으로써 위변조 검출 기법의 발전에 공헌할 것으로 기대된다.

Today, voice files are easily created through a variety of methods, such as phone calls, voice messages, and voice recordings, and are increasingly being submitted as evidence in court. However, due to the nature of digital files, they can be forged and altered, so there is a risk that they can be abused by individuals or organizations with malicious purposes. Accordingly, the importance of forgery detection techniques to ensure the integrity and reliability of voice files is increasing. The purpose of this paper is to examine limitations and suggest development directions through analysis of domestic research on voice file forgery detection techniques. Voice file forgery detection techniques are largely classified into frequency analysis, metadata and file structure analysis, and artificial intelligence utilization. These three detection techniques have limitations in detecting voice file forgery in the case of precise editing. Therefore, this study suggests the need to develop a new algorithm and build a data set to build a deep learning model for detecting forgery of voice files. Additionally, this study suggests the need for a deep learning-based authentication system to prove the integrity of forged voice files before they are submitted as evidence in court. This study is expected to contribute to the development of forgery detection techniques by analyzing the limitations of voice file forgery detection techniques and suggesting development directions.

We propose a speech waveform forgery detection algorithm based on the flatness of frequency distribution. We devise a new measure of flatness which emphasizes the local change of the frequency distribution. Our measure calculates the sum of the differences between the energies of neighboring frequency bands. We compare the proposed measure with conventional flatness measures using a set of a large amount of test sounds. We also compare- the proposed method with conventional detection algorithms based on spectral distances. The results show that the proposed method gives lower equal error rate for the test set compared to the conventional methods.