Cathay Financial Holdings (Cathay FHC) has unveiled pilot deployments of AI-powered “digital colleagues” for project management, technology governance and legal contract review, marking a new phase in the Taiwanese financial group’s digital transformation. Announced at its 2026 Cathay Technology Conference, the initiatives build on the company’s GAIA 2.0 generative AI framework and reflect a broader strategy to integrate governed AI agents into enterprise workflows while exploring stablecoins, AI-agent payments and autonomous finance.
From Generative AI to Agentic AI
The pilots build on GAIA 2.0, Cathay FHC’s generative AI technology framework introduced in 2025. Unlike conventional AI assistants that primarily respond to prompts, agentic AI systems can use designated tools and enterprise information to execute multistep tasks within defined operating boundaries.
Cathay’s initial digital colleagues target project management, technology governance and legal contract review. The company says the agents can autonomously access enterprise knowledge and tools, while critical decisions and exceptions remain subject to human review.
The distinction is important in financial services, where automated workflows may interact with confidential information, regulated processes and business-critical systems. Cathay says its controls include identity and access management, enterprise knowledge permissions, system integration safeguards, cybersecurity monitoring and audit trails.
These controls will be central to determining whether AI agents can move beyond limited pilots into broader financial operations. The announcement does not specify when the digital colleagues will reach production at scale or quantify the efficiency gains achieved so far.
An Agent-First Approach to IT Modernization
Cathay is also extending its existing Cloud First infrastructure strategy toward what it calls an Agent First approach. The objective is to use governed AI workflows and an agent platform to develop capabilities that can be deployed across the group.
One implementation involves an agent developed through the BMad Framework to support requirements management in the software development lifecycle (SDLC). Cathay says the tool has shortened the requirements analysis cycle, although it has not published comparative measurements.
Cathay United Bank has established AI Scrum and AI SDLC teams to support requirements organization, software development, testing and document preparation. These initiatives indicate that the company’s AI strategy extends beyond customer-facing applications into the internal processes used to build and maintain financial technology.
At the conference, speakers including Oliver Jay, OpenAI’s managing director for international, and Harry Lin, Google Cloud Taiwan’s technical general manager, discussed enterprise AI adoption, secure deployment and multi-agent governance. Representatives from Microsoft Taiwan, KPMG and Taiwanese academic institutions also participated.
Cathay reported nearly 7,000 registrations for the conference, with combined online and in-person concurrent viewership exceeding 3,500.
Digital Assets Bring New Governance Questions
The conference also examined virtual-asset custody, stablecoins, AI-agent payments and autonomous finance. These areas could create new ways to initiate transactions and transfer value, but they introduce questions about authorization, fraud prevention, asset security and regulatory compliance.
The Bank for International Settlements has highlighted the growing connections between stablecoins and traditional finance, alongside risks involving financial integrity and financial stability. Its analysis underscores why digital-asset innovation requires appropriate safeguards rather than technology deployment alone.
For AI-agent payments, a central challenge is establishing how an agent receives permission to transact, what limits govern its actions and who remains accountable when an automated decision goes wrong. Cathay’s emphasis on access controls and auditability in its enterprise AI pilots is relevant to that broader governance challenge, although the company has not announced a commercial AI-agent payment product.
A Decade of Digital Transformation
Cathay FHC began its digital transformation in 2016 under Vice Chairman Joseph Tsai. Tsai cited Cathay United Bank’s CUBE credit card, which consolidated multiple cards and allowed customers to switch benefits, as an example of technology designed around customer choice.
The group now plans to continue developing AI, cybersecurity governance, financial technology talent and asset and wealth management capabilities.
The strategic question is whether Cathay can translate its agentic AI experiments into secure, repeatable services across its businesses. Its digital colleagues and SDLC initiatives provide early examples of how the group intends to proceed, while stablecoins and autonomous finance remain areas for further exploration.
For financial institutions evaluating similar approaches, Cathay’s announcement highlights a key implementation principle: autonomous capabilities must develop alongside clear operating boundaries, human oversight and controls that support trust.
Market Landscape
Financial institutions are exploring agentic AI to automate internal processes, assist employees and connect enterprise knowledge with operational tools. Compared with conventional generative AI assistants, agentic systems can execute multistep tasks, making permissions, monitoring and accountability especially important.
At the same time, stablecoins and tokenized financial infrastructure are creating new questions around payment design, financial integrity and regulatory oversight. The BIS has emphasized that these developments require safeguards suited to their risks.
Cathay FHC’s approach brings these trends together through enterprise AI pilots, software development automation and digital-asset research. The company’s next milestones will be the measurable results of its pilots, broader deployment plans and evidence that governance can scale alongside automation.
Top Insights
- Cathay FHC has introduced AI-agent digital colleagues for project management, technology governance and legal contract review, with human oversight of critical decisions.
- GAIA 2.0 provides the foundation for Cathay’s shift toward agentic AI, while its Agent First strategy targets group-wide IT modernization.
- Cathay United Bank is using AI-supported teams and an SDLC agent to assist software requirements, development, testing and documentation.
- The 2026 conference explored stablecoins, virtual-asset custody and AI-agent payments alongside anti-money laundering, cybersecurity and regulatory compliance.
- Cathay has not disclosed detailed productivity metrics or a timeline for scaling its digital colleagues across the financial group.
Get in touch with our fintech expert





