LG headquarters in Yeouido, western Seoul (Yonhap)
LG Group said Wednesday it is launching a new artificial intelligence ethics group later this year with an aim to build more responsible AI systems and their deployment across all LG companies.
Ahead of its official launch, LG AI Research, the group's AI research hub, recently set up a task force to be charged with educating LG employees to better understand AI ethics and evaluating ethical risks related to privacy, bias and fairness before product development.
The upcoming ethics group will be joined by all 10 LG affiliates, including LG Electronics, LG Display, LG Innotek, LG Chem, LG Energy Solution, LG Household & Healthcare, LG Uplus, LG HelloVision, LG CNS and LG Corp.
LG plans to develop more detailed AI ethics guidelines across all affiliates next year based on discussions made through the group.
LG also unveiled five key principles for ethical use of AI – humanity, fairness, safety, accountability and transparency – that will help avert harm, discrimination and other problems with AI development.
In collaboration with LG companies, LG AI Research said it has continued working on risk assessment tools to mitigate ethical issues in AI-powered products and services.
For instance, it is developing a hate speech detection technology that analyzes specific words, even the sentiment of sentences, to address concerns over discrimination, bias and other ethical issues.
Also in development is LG's own Explainable AI or XAI, which helps humans better understand AI systems’ decisions or predictions. The technology is considered essential for supporting human decision-making in areas like medicine, law and finance.
“Evaluating AI systems will depend on how humans perceive and use them,” said Bae Kyung-hoon, head of LG AI Research. “We will elevate customer values by continuing research on the co-existence of humans and AI.”
By Lee Ji-yoon (firstname.lastname@example.org