Wednesday, November 4, 2026

LG AI Research uses AI to identify copyright and privacy risks in AI training data

Input
2026-11-04 10:00:00
Updated
2026-11-04 10:00:00
LG AI Research received the Presidential Council on Intellectual Property Chairperson’s Award for its self-developed AI compliance agent, EXAONE Nexus, at the Copyright Technology Awards at the 2026 International Copyright Technology Conference (ICOTEC).

[Financial News] LG AI Research has been recognized for its technological capabilities after developing an AI agent that analyzes copyright and privacy risks in AI training data. The aim is to reduce the time and cost of data review and address legal risks that may arise during AI development. 
Risk-assessment results for about 20,000 datasets released

LG AI Research said on the 4th that it received the Presidential Council on Intellectual Property Chairperson’s Award for its self-developed AI compliance agent, EXAONE Nexus, at the Copyright Technology Awards at the 2026 International Copyright Technology Conference (ICOTEC), hosted by the Ministry of Culture, Sports and Tourism. The Presidential Council on Intellectual Property is an agency under the Prime Minister.
EXAONE Nexus is a system that automatically analyzes copyright, licensing and privacy risks in large-scale AI training data. It is designed to examine vast datasets that are difficult for people to review manually and assess legal risks in the AI development process.
According to LG AI Research, when analyzing a single dataset, the system was 45 times faster and 26% more accurate than legal experts. The institute said the analysis cost was about one-thousandth of the cost of hiring legal experts.
Controversy over copyright infringement in AI training data is raising issues of public trust in the technology and companies’ legal responsibility. LG AI Research said it developed EXAONE Nexus in response and is also building its own compliance framework reflecting regulations in major jurisdictions, including South Korea’s AI Basic Act and the European Union’s (EU) AI bill.
To verify the legal validity of its analysis results, the institute worked with around 40 lawyers from major law firms in South Korea and abroad, including the United States. It released to academia and the research community the risk-assessment results for about 20,000 open datasets used in empirical analysis.
The institute uses EXAONE Nexus as a mandatory validation step for every AI model developed internally. It said the system examines data-related risks throughout the AI development process, from data collection and curation to model development and service deployment. It plans to expand supply to overseas markets with demand for compliance solutions, including among global big tech companies and law firms.
Hwayoung Lee, head of AI Business Development at LG AI Research, said, “Securing public trust amid the rapid advancement of AI technology is a key competitive advantage for companies,” adding, “We will continue to introduce AI technologies that combine high reliability with outstanding performance.”
1,131 patent applications filed... AI ethics education also expanding

LG AI Research is also working to protect intellectual property rights and foster a responsible AI ecosystem. It has established an intelligent system for reporting research achievements as papers or inventions and has filed 1,131 patent applications in South Korea and abroad since its founding in 2020. Of these, 537 were filed domestically and 594 overseas.
Yoo Gyeong-jae, IP Strategy Lead at LG AI Research, who contributed to building this system, received the Prime Minister’s Award for Science and ICT at the ICT Patent Management Awards on the 2nd.
The institute said that since its launch in 2020, it has participated at least 30 times in global discussions on AI ethics and governance. As a representative of South Korean companies, it has attended events including the UNESCO Global Forum on the Ethics of AI, the AI Impact Summit and the Paris AI Action Summit, where it has presented the role of businesses in sustainable AI development. Last July, it launched an online open course (MOOC) on AI ethics education with UNESCO, which has attracted about 50,000 learners worldwide to date.
These activities also align with LG’s management approach of making compliance a foundation for corporate growth. Koo Kwang-mo, CEO of LG Corp., has emphasized that compliance should be seen as essential infrastructure for a company’s growth and development, and that related systems should be upgraded to reflect changes in the times and society.
[email protected] Yoon Hong-jip Reporter