Proprietary Project
Samsung Electronics is developing a generative artificial intelligence (AI) service for employees. This strategy aims to boost productivity and work efficiency through customized AI developed in-house as security issues such as a leakage of confidential documents due to AI are on the uptick.
According to sources in the Korean business community on June 12, the Device Solution (DS) Division in charge of Samsung Electronics’ semiconductor business, is developing its own large language model (LLM) at the GPT-3.5 level or higher within this year. The service will support a total of nine areas -- automated responses to business processes such as purchasing and expenses, searches for specialized knowledge such as processes, design, and manufacturing, summarizing manufacturing and process data, translation, document production, recording and summarizing meeting minutes, market and company analyses, creating and reviewing code, and responding to the voices of customers.
The AI service that Samsung Electronics is developing is expected to begin its basic service as early as this December and provide specialized search services that include the company’s knowledge and data in February next year.
Currently, Samsung Electronics temporarily banned employee use of generative AI in the Device Experience (DX) Division earlier this month. This ban was put out of concerns about internal information leakages from uploading internal emails and entering internal source code in March.
This development is part of the Korean tech giant’s strategy to enhance work efficiency by introducing customized AI developed in-house as security issues such as trade secret leaks due to the spread of external generative AI gain in importance, analysts say.
In the meantime, President Kyung Kye-hyun has been showing significant interest in AI. His ambition is to explosively increase demand for semiconductors after smartphones through a generative AI craze.
Some analysts believe that Kyung’s stance to invest as planned without artificially cutting chip production despite a recent downturn in the semiconductor market may be closely connected to preparations for a boom in semiconductors brought about by a generative AI craze.
In fact, the Wall Street Journal (WSJ) reported that demand for graphics processing units (GPUs) that can power AI has surged due to a generative AI craze, causing shortages of high-end semiconductors.