Liner, a domestic AI search specialist, has embarked on improving its system and expanding its workforce with the goal of providing a safe artificial intelligence (AI) service. Liner announced on the 5th that it will strengthen its safety measures within its service to address the problem of harmful search terms that have increased with the recent surge in users.
|
▲[Korean Today] Liner homepage main screen. © Reporter Byun A-rong |
Liner is a search engine specialized in paper searches and assignments, and in the beginning, there were almost no searches for harmful keywords, but as the number of users increased, the number of cases of searching for sensitive keywords such as ‘depression’ and ‘suicide’ increased. This has already emerged as a major problem in global AI services. For example, last year, Character.AI was embroiled in controversy because a certain chatbot continuously induced negative conversations with users.
In response, Liner has introduced a system that displays positive messages and cute characters when searching for harmful keywords. This measure is part of an effort to help alleviate negative emotions. In addition, the company is working on improving technology to apply the safety guardrails of the Large Language Model (LLM) more strictly.
Recruitment to enhance the safety of the service is also expanding. Liner recently posted a job posting for a ‘Security Compliance Engineer’ and is currently recruiting personnel to take charge of service security and compliance. A Liner official said, “We will actively secure professional personnel to continuously improve safety devices.”
Currently, Liner is hiring more than 20 additional employees, including developers, and is on the verge of surpassing 100 total employees. It remains to be seen whether Liner’s efforts to develop good AI will become a positive example within the industry.
© The Korean Today – All Rights Reserved