SIAgentNews.com
← All news

Softcamp unveils security tools to manage financial AI agent permissions

The company is launching three new solutions to help banks monitor what data AI agents can access and prevent data leaks.

· First reported by 머니투데이

12thDate of FISCON 2026 event

The short version

  • SHIELD AI Gateway verifies agent identities and records every request made to large language models or business tools.
  • SHIELD Gate uses Remote Browser Isolation to keep employee access to external AI separate from internal business terminals.
  • XSCAN Secure Asset uses a Software Bill of Materials to track vulnerabilities across financial service components.
  • The tools are designed to help banks audit which specific data an AI agent accessed and under whose authority.

Why it matters

As banks move toward using autonomous AI agents, they need a way to ensure these bots don't access sensitive data they shouldn't see. These tools provide a framework for oversight, making it safer for large organizations to integrate AI into daily operations.

New security tools for banks

Softcamp has announced its participation in the Financial Information Security Conference, known as FISCON 2026. The event is scheduled to take place on the 12th at the Conrad Seoul in Yeouido, Seoul. During this conference, the company plans to showcase several new security solutions tailored for the financial sector.

The primary goal of these new products is to help financial institutions manage the risks associated with generative AI. As these organizations begin to incorporate AI into their daily workflows, they face unique challenges regarding data privacy and system access. Softcamp aims to provide the infrastructure needed to balance innovation with safety.

The company will be joined by its subsidiary, Redpen Software, to exhibit a suite of tools. These solutions cover a broad range of needs, including managing external AI access and monitoring software supply chains. The focus is on creating a secure environment where AI can operate without compromising sensitive information.

Controlling AI agent permissions

One of the core products being unveiled is the SHIELD AI Gateway. This tool is designed to manage the specific process by which AI agents interact with large language models (LLMs) or other business tools. It acts as a gatekeeper that verifies the identity and permissions of an agent before allowing it to proceed with a task.

When an AI agent attempts to access data or invoke a specific tool, the gateway checks if that action is authorized. If the request is approved, the system relays the call and maintains a detailed history of the interaction. This allows financial institutions to see exactly which agent accessed which system and under whose specific authority it did so.

These logs are particularly important for internal controls and auditing. By maintaining a clear record of every action, the company helps banks meet regulatory requirements. It ensures that no AI agent can perform unauthorized tasks or access private data without a traceable permission trail.

Preventing data leaks

To address the risks of using external generative AI services, Softcamp is introducing SHIELD Gate. This solution utilizes Remote Browser Isolation (RBI) technology to create a barrier between employees and external AI platforms. RBI works by separating the browser session from the actual business terminal used by the worker.

By isolating the connection, the company aims to reduce the risk of business data leakage. Employees can still interact with external AI tools, but the sensitive data on their local machines remains protected from the external service. This creates a safer way for staff to use popular AI models without moving private files into the cloud.

The objective is to provide a controlled environment for human-AI interaction. It allows for the benefits of external AI while maintaining a strict perimeter around the company's internal data. This is a critical step for industries that handle highly regulated and private information.

Tracking software vulnerabilities

Redpen Software, a subsidiary of Softcamp, will also present a tool called XSCAN Secure Asset. This product is based on the Software Bill of Materials (SBOM), which is essentially a list of every component included in a piece of software. It helps financial services understand the full makeup of the software they use.

The tool identifies the various components within a financial service and monitors them for security flaws. When a new vulnerability is discovered in a common component, XSCAN can track which specific assets are affected. This allows IT teams to respond much faster to security threats.

Furthermore, the system tracks the history of remediation for these vulnerabilities. This provides a clear picture of what has been fixed and what still needs attention. It helps organizations maintain a more transparent and secure software supply chain in a complex digital landscape.

The path toward secure AI

Bae Hwan-guk, the CEO of Softcamp, emphasized the necessity of these tools for the financial sector. He noted that for banks to use AI effectively, they must be able to verify what data is being sent externally. He also highlighted the need to know what tasks AI agents are performing and who authorized those specific actions.

The company's strategy is to provide comprehensive security that covers multiple layers of the AI lifecycle. This includes everything from the initial access to external models to the management of internal agent permissions. By addressing these different layers, Softcamp hopes to provide a complete security package for the financial industry.

As the conference approaches, the focus remains on providing practical solutions for the field. The goal is to move beyond theory and offer tools that can be deployed immediately. These solutions aim to give financial institutions the confidence to adopt AI while keeping their data and systems secure from unauthorized access.

What to watch

  • How financial institutions will adopt SHIELD AI Gateway for auditing.
  • The adoption rate of Remote Browser Isolation in the banking sector.

Sources and links

First reported by 머니투데이. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.

Liked it? Say why

Add a line for other readers.

No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.

More AI news