Establishing Cyber Threat Response Technology for Both Companies
Red Team Collaboration and Development of Safety Evaluation Standards
Aim Intelligence, a startup specializing in generative artificial intelligence (AI) security technology, is partnering with KT to collaborate on responding to cyber threats and promoting the distribution of safe AI models.
On the 3rd, the two companies announced that they had signed a Memorandum of Understanding (MOU) for joint research on security and safety verification systems for AI-based models and the establishment of threat response technologies.
The two companies agreed to mutually cooperate in areas including ▲joint research on security and safety verification system technologies for generative AI-based models ▲development of model defense and reinforcement technologies to address identified vulnerabilities ▲responding to cybersecurity threats and promoting the distribution of safe generative AI models.
Aim Intelligence stated that to promote the distribution of safe AI models, it will provide generative AI safety diagnostics (Red Teaming) and integrate automatic vulnerability detection and blocking/defense technologies (Guardrail).
The Aim Intelligence Red Team is composed of winners from the generative AI Red Team Challenge hosted by the government last April.
Generative AI Red Teaming refers to structured testing activities that use adversarial methods in a controlled environment to identify defects and vulnerabilities such as harmful or discriminatory outcomes and system misuse in AI systems. AI companies both domestically and internationally, including OpenAI, Google, and Naver Cloud, have formed and operate their own AI Red Teams, highlighting this as a key method for securing AI safety and reliability.
Through this agreement, the two companies will carry out various activities for generative AI safety and security. Aim Intelligence and KT Red Teams will collaborate to jointly develop benchmarks that serve as safety evaluation standards for generative AI models.
Hwang Tae-seon, Chief Information Security Officer (CISO) of KT, said, "We will do our best to cooperate to promote the distribution of safe generative AI models." Yoo Sang-yoon, CEO of Aim Intelligence, stated, "We will work together with KT to provide safe generative AI services that anyone can trust and use."
Aim Intelligence has been researching and developing core technologies for generative AI security after going through Naver’s investment company D2 Startup Factory. CEO Yoo plans to advance security technologies for the upcoming Artificial General Intelligence (AGI).
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
