SIAgentNews.com
← All news
DevelopingPolicy

Singapore firm iCOMPASS launches KY-AI for financial AI governance

The new platform helps banks and financial firms track, test, and oversee autonomous AI systems to meet upcoming regulatory requirements.

· First reported by Macau Business

Isometric diorama: picture for the story 'Singapore firm iCOMPASS launches KY-AI for financial AI governance'
AI-made isometric diorama for this story. No real people or products are shown.
7 October 2027Date new MAS AI guidelines take effect

The short version

  • KY-AI helps firms manage both internally developed and third-party AI systems throughout their entire lifecycle.
  • The platform follows new AI Risk Management Guidelines from the Monetary Authority of Singapore (MAS).
  • It replaces fragmented spreadsheets with a structured process for approvals, testing evidence, and risk tracking.
  • The tool is designed to ensure human accountability remains central as AI becomes more autonomous.

Why it matters

As AI moves from simple tools to autonomous agents, financial firms need a way to prove they are controlling these systems. This platform helps them meet legal standards without slowing down their adoption of new technology.

A new standard for AI oversight

Financial institutions are facing a growing challenge as artificial intelligence shifts from basic productivity tools to autonomous systems. These new systems can make recommendations and execute tasks independently, making it harder for companies to understand what the technology is doing. Organizations must now figure out how these systems behave and who is responsible when they make mistakes.

To address this, the Singapore-based regulatory technology firm iCOMPASS launched a platform called KY-AI. The name stands for Know Your Artificial Intelligence, drawing a parallel to the long-standing Know Your Customer requirements in banking. The goal is to provide visibility into the AI systems operating within a company and the decisions they influence.

This launch comes in response to the Monetary Authority of Singapore (MAS) issuing its AI Risk Management Guidelines on 7 October 2026. These rules set out the supervisory expectations for how the financial sector should manage risks associated with AI. The guidelines are scheduled to take effect on 7 October 2027, creating a deadline for firms to modernize their oversight.

Moving beyond one-time checks

Many companies currently treat AI risk assessment as a one-time task performed before a tool is deployed. KY-AI is designed to change this by supporting documented oversight throughout the entire lifecycle of an AI system. It moves away from periodic reviews toward a continuous process of monitoring and evidence gathering.

The platform is structured around four specific stages: Know, Approve, Test, and Watch. These steps allow organizations to register their AI use cases and identify the specific owners responsible for them. It also helps teams assess dependencies and document the approvals needed before a system goes live.

By using a structured process, the platform helps firms track material changes that might require a new review. This is particularly important for autonomous agents that process sensitive information or analyze complex data across interconnected systems. The system aims to replace fragmented spreadsheets and disconnected approvals with a single, repeatable workflow.

Managing third-party and internal tools

A major hurdle for financial institutions is the mix of technology they use. Companies often rely on a combination of internally developed software and third-party tools provided by outside vendors. KY-AI is built to manage both types of systems, recognizing that a firm remains accountable for AI risks even if they did not build the underlying technology.

The platform connects different departments, including business owners, risk teams, and compliance functions. It allows these groups to record decisions regarding what information a tool can access and when a human must review the output. This ensures that technical reviewers and compliance officers can work together in a unified environment.

The developer of KY-AI also created the ULTRA compliance operating system, which was featured in the MAS Pathfinder programme. By extending this existing technology into AI lifecycle governance, iCOMPASS aims to make oversight easier to evidence. The platform seeks to provide clear accountability while maintaining human approval at every critical stage of the process.

Balancing innovation and safety

The goal of the new platform is to ensure that AI governance does not become a burden that prevents companies from using new technology. Raymond Moh, the CEO of iCOMPASS, stated that the goal is to help organizations adopt AI with confidence while using proper safeguards. The platform is intended to provide the necessary structure for safe adoption rather than just creating more paperwork.

By providing a way to translate complex regulatory obligations into measurable outcomes, the tool helps firms stay compliant while remaining competitive. It allows them to identify which AI systems are operating within their organization and what specific decisions those systems can influence. This level of detail is becoming essential as AI becomes more embedded in daily business operations.

iCOMPASS will demonstrate the KY-AI platform at the Singapore FinTech Festival 2026, which runs from 18 to 20 November. The solution will also be displayed at the MAS Pavilion during the Future of Finance showcase. Visitors will be able to see how the platform handles the practicalities of AI governance in a regulated environment.

What to watch

  • How other global regulators might adopt similar 'Know Your AI' frameworks.
  • Whether other RegTech firms will launch competing platforms to meet the 2027 deadline.

Try it

Check the Monetary Authority of Singapore (MAS) website for the full text of the AI Risk Management Guidelines issued in October 2026.

Sources and links

First reported by Macau Business. 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