According to industry sources, Samsung Electronics is reportedly developing an in-house artificial intelligence service for internal use, responding to security concerns surrounding OpenAI’s advanced chatbot, ChatGPT. Co-CEO Kyung Kye-hyun, responsible for the tech giant’s chip business, revealed the plans during a recent lecture attended by Samsung executives and employees, emphasizing the need for a customized approach.
The upcoming AI service, as mentioned by Kyung, will employ a large language model to support various tasks within the chip business division, including knowledge search, translation, and other administrative functions. The company aims to launch a basic version of the service in December, with a more integrated offering scheduled for February the following year. The AI-based system is expected to enhance nine key areas, such as automatic response for product purchases, cost management, data analysis for chip production, and transcription of internal meetings.
During the lecture, Kyung also expressed Samsung’s commitment to contributing to the AI ecosystem by leveraging their exceptional technology. He discussed potential future directions for AI-integrated solutions, such as wide bandwidth, high power efficiency, and an integrated architecture spanning smartphones to data centers.
The decision to develop an in-house AI service comes in light of incidents where Samsung’s DS division encountered misuse cases involving sensitive company information uploaded to external generative AI tools like ChatGPT. As a result, the company has restricted the usage of such tools and is now focusing on customized AI solutions provided by local specialized firms.
During a separate lecture at Yonsei University, the Samsung CEO praised ChatGPT as an impressive tool and mentioned that the company plans to incorporate the advanced AI technology in some capacity from the next year, while ensuring appropriate security measures are in place.