[AI Accountability Report on AI Ethics EP.1] Inclusive AI - From attending the AI Action Summit to Publishing AI Accountability Report
LG AI Research continues its AI ethics initiatives to ensure that everyone can benefit from AI.
From the 10th (local time), LG AI Research participated in the AI Action Summit held over two days at the Grand Palais in Paris, France, presenting its AI ethics achievements. The AI Action Summit brings together heads of state, business leaders, and academic figures from over 100 countries to discuss AI’s potential and risks in a balanced manner, including AI for public good and its impact on the environment.
In particular, as part of the Paris AI Action Summit, LG AI Research, in collaboration with UNESCO, hosted a session titled "Preparing AI for the World." This session focused on discussing international cooperation strategies for inclusive and responsible AI governance while sharing best practices.
The session was attended by Gabriela Ramos, Assistant Director-General for Social and Human Sciences at UNESCO, along with ministers from various countries, corporate representatives, and policymakers, totaling approximately 300 attendees.
During the session, the two organizations shared interim results of their jointly developed global AI ethics online education program (MOOC: Massive Open Online Course). Furthermore, both parties committed to initiatives that promote the practice of AI ethics, including the launch of a global competition for best practices in AI ethics, featuring contributions from international AI ethics experts.
Alongside its participation in the AI Action Summit, LG AI Research also released the LG AI Ethics Accountability Report, which details the organization’s progress in AI ethics. Notably, the report highlights LG’s efforts in realizing Responsible AI and Inclusive AI, underscoring its commitment to creating a world where all members of society can benefit from AI technology.
heck out some of the AI ethics initiatives and achievements LG AI Research has been working hard on and bringing to life over the past year!
▶ View the LG AI Accountability Report on AI Ethics in detail
[AI Accountability Report on AI Ethics EP.1] “Inclusive AI For Everyone’s Benefit”
1. Fostering Openness by “Sharing Models to Improve Accessibility to AI Technology
First, we at LG AI Research Lab believe that AI technology should be more accessible, so we released EXAONE 3.0 and the upgraded EXAONE 3.5 model as open source models. By doing so, we hope that this is the start of “AI for everyone” as researchers around the world can freely utilize our model for research purposes and contribute to the revitalization of the AI ecosystem.
Image 1. EXAONE Model Release Milestone and Red Teaming Test Results of EXAONE 3.5
In addition to the high performance of EXAONE 3.0, which was released in August 2024, and EXAONE 3.5, which was released in December 2024, we have worked hard to ensure that our models are “ethical models.” We have conducted a detailed evaluation of the ethics and security of our models through a red teaming process using internal and external datasets, and have further made our evaluation results transparent so that researchers around the world can gain insights from them and continue their work on ethical models.
2. Education by “Providing Quality AI Education to Alleviate Socioeconomic Disparities”
One of the most important parts of “AI for everyone” is quality AI education that is accessible to all. We run age-specific programs to expand access to quality AI education. We provide hands-on education that combines theory and practice through several different programs, including LG Discovery Lab for middle and high school students, LG Aimers for college students, and LG AI Academy for working professionals.
Image 2. Customized AI Education Programs Provided by LG
3. Partnership via “Advancing the UN Sustainable Development Goals through Global Collaboration”
As AI technology rapidly advances and its applications expand, AI ethics has become a crucial challenge that requires global cooperation beyond individual nations. That's why we're also working with a wide range of external experts to create macro-level change beyond our own activities. Together with UNESCO, we're developing a Massive Open Online Course (MOOC) to raise awareness of ethical AI development and use among AI professionals and policymakers around the world.
We also put our money where our mouth is to foster real change in AI ethics. As previously mentioned, we are collaborating with UNESCO at the AI Action Summit to host a call for best practices involving international AI ethics experts. The selected cases will be incorporated into MOOC educational content and shared widely as part of our ongoing efforts.