Nakao Adopts Joint Declaration to Prevent Malicious Election Deepfakes

The Korea Internet Corporations Association announced on the 8th that its member companies Naver, Kakao, and SK Communications have adopted a joint declaration to prevent the malicious use of deepfakes in elections.

The joint declaration is part of the activities of a voluntary consultative body formed by major IT companies to respond to malicious election deepfakes. It was established to set a close cooperation framework among participating companies during the 22nd National Assembly general election period.

The main contents include detection and prompt action against malicious election deepfakes, disclosure of response policies, activation of exchanges to prevent spread, and raising public awareness. Since the services provided by each company vary, additional measures tailored to their respective characteristics are also being prepared.

Naver is promoting the adoption of global technical standards that allow verification of content source information and is researching the acquisition of generated content detection technology as a mid- to long-term goal. Kakao is preparing to introduce technology that attaches invisible watermarks to its generative AI image creation model "Karlo." SK Communications prevents the production of election-related deepfakes by restricting search terms related to major politicians' names when generating images using AI.

The Korea Internet Corporations Association stated, "Big tech companies have announced policies to respond to deceptive AI election content through the Munich Agreement," and added, "We will participate in the voluntary consultative body activities with domestic companies that adopted this declaration according to the technical agreement."

This content was produced with the assistance of AI translation services.