Moves by corporations to adopt generative artificial intelligence (AI) in their work are accelerating, but in the field its use is often limited by security and internal control issues. In particular, the stricter the regulations, as with public institutions and the financial sector, the more cases there are in which AI adoption stops at the pilot stage. ITCEN CloIT released an enterprise artificial intelligence (AI) platform to address these problems.
On the 27th, ITCEN CloIT held a press briefing at the Grand Intercontinental Seoul Parnas in Gangnam-gu, Seoul, and said it will launch "AgentGo 2026," aimed at organizations with strict security regulations such as the public and financial sectors.
ITCEN CloIT, a cloud and AI subsidiary of ITCEN Group, has developed cloud, data, security and AI businesses based on its experience building IT systems for the public sector, finance and large corporations. The company said that although generative AI can improve work efficiency, there are limits to applying it to actual work due to concerns that internal data could leak outside and the burden of complying with regulations.
Kim Woo-seong, CEO of ITCEN CloIT, said at the briefing, "Until now, corporations have focused on how much AI they adopt, but now it has become more important how well they use AI," adding, "AgentGo is a platform focused on helping corporations reduce trial and error and quickly settle AI in line with their respective work levels and system environments."
According to market research firm IDC, by 2027 more than 40% of enterprise applications will be enhanced through automation using AI agents. However, ITCEN CloIT said that many AI projects are stopping at the proof-of-concept (PoC) stage due to security concerns and integration issues with existing systems.
Vice President Cho Sang-cheol of ITCEN CloIT said, "AgentGo is not just a tool for building a chatbot that answers questions, but a platform for a company to manage AI that performs real work," adding, "The core is a structure that centrally controls multiple AIs and has them collaborate."
The newly unveiled "AgentGo 2026" is designed not only to run AI directly on a company's internal servers, but also to be used in an internal, dedicated network environment completely separated from the external internet. It is structured so that corporations can use AI only within their in-house systems without sending data outside.
Cho said, "Recently, corporations value a structure in which data does not leave the company more than AI performance," adding, "Especially in finance and the public sector, it is often difficult to use external servers, so demand to operate AI within internal systems is rising again."
The platform adopts a method of managing and controlling multiple AI programs (agents) in one place. It defines the types of AI and access scopes available to employees or by department, and allows real-time checks of what tasks the AI has performed.
Cho said, "If you don't define who uses which AI and how far they can access, AI can actually become dangerous," adding, "AgentGo is designed to automatically block cases where an unauthorized user requests sensitive information or issues commands that deviate from the rules."
The approach to work automation is also different from before. Unlike RPA (robotic process automation), where a person directly clicks through screens to handle repetitive tasks, the explanation is that it implements "action-centric automation," in which AI checks company rules and systems together and then proceeds through to actual task processing.
For example, if you connect an AI that checks in-house travel expense rules with an external airline ticket reservation system, the company said it can first verify whether there is a rule violation and, if there is no issue, automatically proceed through the actual booking.
ITCEN CloIT offers AgentGo 2026 in three forms depending on the customer environment: a chatbot type focused on in-house document search; a standard type that automates tasks by linking with external systems; and a high-security type that operates only on an internal network without external connections. For organizations where security is especially critical, it also unveiled an all-in-one product that uses dedicated server equipment.
Kim said, "We are currently conducting a PoC related to a pharmaceutical company's new drug development process, and discussions on adoption are continuing in finance and the public sector," adding, "In line with the government's trend of expanding the AI budget, we expect examples of adoption to grow quickly in the public sector as well."
There were also remarks about changes AI adoption will have on organizational structures. Kim said, "If AI agents take on repetitive tasks, additional hiring for simple tasks could decrease," adding, "Instead, there is a high possibility that the work structure will change so that junior staff achieve higher levels of performance by using AI." Kim added, "AI is now less a technology than a colleague who works with the company," emphasizing, "AgentGo's goal is to manage AI so it does not stray from the organization's rules while ensuring it leads to actual work results."