Search

KR-102961397-B1 - Generative AI portal service method and service system thereof

KR102961397B1KR 102961397 B1KR102961397 B1KR 102961397B1KR-102961397-B1

Abstract

The present invention relates to a generative AI portal service method and system. In addition, the present invention provides a processor that registers information about at least one generative AI operating server and links with security equipment within the enterprise to block entry to the generative AI operating server on the enterprise network, a filtering step that receives a query message from a user terminal within the enterprise on behalf of the generative AI operating server whose entry is blocked and filters the query message according to a predefined security policy, and a query response step that transmits the filtered query message to the generative AI operating server to receive a response message to the query and provides the received response message to the user terminal, thereby preventing the leakage of internal information of the enterprise and providing a safe usage environment for generative AI.

Inventors

  • 양우석
  • 이윤석

Assignees

  • 주식회사 수산아이앤티

Dates

Publication Date
20260511
Application Date
20240530

Claims (10)

  1. A registration step in which a processor registers information about at least one generative AI operation server and links with security equipment within the enterprise to block entry to the generative AI operation server on the enterprise network; A filtering step that receives query messages from user terminals within the enterprise on behalf of a generative AI operation server whose access is blocked, and filters the query messages according to a predefined security policy; A question-response step of transmitting a filtered query message to a generative AI operation server to receive a response message for the query and providing the received response message to a user terminal; and A generative AI portal service method characterized by including: a providing step of storing information regarding the content of a query message transmitted to a generative AI operation server and the usage status of the generative AI operation server, and generating a security report based on the stored information and providing it to an administrator terminal.
  2. In Article 1, The above registration step is, A generative AI portal service method characterized by redirecting to a URL where a query message is received on behalf of the generative AI operation server when a request to enter the generative AI operation server is made.
  3. In Article 1, The above filtering step is, A generative AI portal service method characterized by providing a selection interface that allows selecting at least one of the pre-registered generative AI operating servers before receiving a query message.
  4. In Article 1, The above filtering step is, A generative AI portal service method characterized by defining information, including internal corporate information and development code, for which information leakage must be restricted, as a security policy, and the said defined security policy being updated or modified at regular or irregular intervals.
  5. delete
  6. processor; and A memory for storing one or more instructions executable by the above processor; and The above processor, by executing the above one or more instructions, Register information on at least one generative AI operation server and integrate with corporate security equipment to block access to the generative AI operation server on the corporate network, and It receives query messages from user terminals within the enterprise on behalf of a generative AI operation server with blocked access, and filters query messages according to predefined security policies, Filtered query messages are transmitted to a generative AI operation server to receive response messages for the queries, and the received response messages are provided to user terminals. A generative AI portal service system characterized by storing information regarding the content of query messages transmitted to a generative AI operation server and the usage status of the generative AI operation server, and generating a security report based on the stored information and providing it to an administrator terminal.
  7. In Paragraph 6, The above security equipment is, A generative AI portal service system characterized by redirecting to a URL where a query message is received on behalf of the generative AI operation server upon a request to enter the generative AI operation server.
  8. In Paragraph 6, The above processor is, A generative AI portal service system characterized by providing a selection interface that allows selecting at least one of the pre-registered generative AI operating servers before receiving a query message.
  9. In Paragraph 6, The above processor is, A generative AI portal service system characterized by defining information, including corporate internal information and development code, for which information leakage must be restricted as security policies, and wherein the defined security policies are updated or modified at regular intervals or irregularly.
  10. delete

Description

Generative AI portal service method and service system thereof The present invention relates to a generative AI portal service method and system. Generally, many companies recommend the use of generative AI, such as ChatGPT, as it provides significant assistance to business operations. As such, with the active use of generative AI, big tech companies such as Google, Microsoft, and Naver, in addition to ChatGPT, are announcing their own developed generative AI or striving to expand the market. However, as companies utilize generative AI, there is a problem that information leakage regarding internal corporate information may occur, and security threats resulting from this increase. In other words, since generative AI understands input commands based on data learning using artificial neural networks and generates and provides new content such as text, images, audio, and video in accordance with the user's intent, information leakage regarding internal data may occur as the AI learns internal information and development code requested from the generative AI and provides them to other users. Therefore, while companies are attempting to prevent security threats in advance by introducing on-premises Network DLP (Data Loss Prevention) equipment, there is a problem in that it is virtually impossible to prevent all security threats with on-premises Network DLP equipment due to the rapid increase in generative AI protocols and the significant internal changes occurring. Furthermore, even if one attempts to apply on-premise Network DLP equipment to some generative AIs, it is difficult to do so easily due to reduced costs, performance, and stability. Meanwhile, the technology forming the background of the present invention is disclosed in Published Patent Application No. 10-2018-0081383 (July 16, 2018) and Registered Patent Application No. 10-2604242 (Registered Nov. 15, 2023). FIG. 1 is a drawing of a generative AI portal service system according to the present invention. FIG. 2 is a block diagram showing the hardware configuration of a generative AI portal service system according to the present invention. FIG. 3 is a block diagram showing the functions performed by the processor of a generative AI portal service system according to the present invention. FIG. 4 is a flowchart of a generative AI portal service method according to the present invention. In the following description, specific details of the invention are provided to provide an overall understanding of the invention, but it will be obvious to those skilled in the art that the invention can be easily practiced without these specific details and by variations thereof. Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the attached FIGS. 1 to 4, focusing on the parts necessary to understand the operation and function according to the present invention. FIG. 1 is a drawing of a generative AI portal service system according to the present invention. Referring to FIG. 1, a generative AI portal service system (100) according to one embodiment of the present invention may be implemented as a server capable of communicating with a generative AI operation server (10) or as a program installed on said server. Here, the generative AI portal service system (100) according to one embodiment of the present invention is positioned between a generative AI operation server (10) and a company and acts as a proxy that monitors and controls inquiries made from the company to the generative AI operation server (10), thereby enabling the generative AI to be utilized in business without information leakage regarding the company's internal information, development code, etc. That is, the generative AI portal service system (100) according to one embodiment of the present invention can minimize security threats caused by generative AI while simultaneously providing a safe usage environment to the company. At this time, the aforementioned generative AI operating server (10) may include ChatGPT, Google Bard, and Microsoft Bing, and may also include a service or server that provides various generative AI models. In addition, the generative AI portal service system (100) according to one embodiment of the present invention may be built within a company on an on-premises basis or built on a communication network outside the company on a cloud basis. In the case of the aforementioned cloud-based system, it is preferable to perform a predetermined service subscription procedure to allocate a service domain or ID, etc., to each company. In addition, it is desirable to use at least one security device (20) already established within the company, such as a Secure Web Gateway (SWG), to block user terminals (30) within the company from entering the generative AI operation server (10) through the company network and the internet network, and to allow querying the generative AI operation server (10) only through the generative A