AI Elections Accord: Global Companies Team Up to Combat Election Deepfakes
AI technology has now become a natural part of our everyday lives, providing us with various innovations we’ve never experienced before. At the same time, we are seeing cases in which such advances in AI are exploited for malicious purposes, most notably through the use of deepfakes. Deepfakes are considered problematic because they not only cause direct harm through fake digital content but also may lead to distrust of digital content and AI technology.
Now, as the world’s “super election year” begins, the threat posed by deepfakes is taking a new turn. With elections held in 76 nations and half the world’s population exercising their voting rights, there has been a growing concern that fake political content generated by deepfakes may influence election results. Accordingly, in an effort to prevent these potential threats and steer the use of AI technology in the right direction, 20 global companies have come together and announced the AI Elections Accord at the Munich Security Conference, the largest conference in the field. The signatories of the accord include Google, Microsoft, Meta, and Amazon, with LG AI Research becoming the only Korean company to join them and pledge its cooperation at a global level.
2024 AI Elections Accord: 20 global companies team up to combat election-related deepfakes
Most notably, the AI Elections Accord is a joint effort by companies that are leading the development of AI technology and directly or indirectly responsible for the distribution of generative AI content. These include big tech companies such as Google, Microsoft, Meta, Amazon and AI companies such as OpenAI, Anthropic, and Stability AI, as well as social media companies like Snap, TikTok, and X, along with the semiconductor company ARM.
Image 1. Signatory Companies of the AI Elections Accord
LG AI Research becomes the only Korean company to pledge action against the misuse of AI alongside other global companies
With the participation of various global companies, the AI Elections Accord plans to establish safety devices for the entire process from the creation to the consumption of deepfake election content. Especially, to minimize election risks posed by AI-generated content, the companies have pledged to invest in developing technology such as invisible watermarking that reveals the origin of content. They also plan to launch campaigns to prevent online platform users from spreading deepfake content that may influence voters or public opinion. The AI Elections Accord is significant in that companies around the globe have come together to make AI technology trustworthy and implement self-regulations for fair elections.
Image 2. Seven Principle Goals of the AI Elections Accord
As the only Korean company in the accord, LG AI Research pledges to take the following actions
By signing the AI Elections Accord, LG AI Research’s endeavor toward AI ethics expands to fighting deepfakes and reinforcing the fairness of elections. We will develop technological solutions for identifying the origin and authenticity of deepfake content and work with other signatory companies to discuss global standards for responding to deepfake content.
Image 3. LG AI Research’s ethical commitments
LG AI Research’s commitment to AI ethics for the past several years has positioned us as a global leader in ethics, making it the only Korean company to sign the AI Elections Accord. After establishing our own AI ethics, we published the Accountability Report on AI Ethics, which outlines our achievements in practicing AI ethics, and in February, attended UNESCO’s Global Forum on the Ethics of AI as the only Korean company to discuss cooperation with global organizations and companies. With the signing of the AI Elections Accord, LG AI Research pledges to work with global companies to prevent deepfakes from affecting elections or fostering political distrust. Additionally, we aim to lead AI ethics efforts globally through various actions.