…
6 November 2026
Introduction
On 2 November 2026, the Securities and Futures Commission (the “SFC”) published its first Large Firm AI Readiness Survey (the “Survey”), covering 20 large financial firms and 78 affiliated licensed corporations that collectively account for over 75% of reported securities transactions in Hong Kong.
The central message is clear: AI is moving beyond standalone productivity tools and becoming embedded in firms’ systems, workflows and decision-making processes. As deployments scales and AI applications become more autonomous, the regulatory focus is correspondingly shifting towards governance, operational resilience and clear human accountability.
The Survey therefore provides an important indication of the SFC’s emerging supervisory expectations as financial institutions move from AI experimentation to scaled deployment.
Key Survey Findings
- AI Is Moving into Core Operations
The Survey shows that AI adoption among large financial institutions has entered a more mature phase. While 95% of respondents expect AI use to increase over the next 12 months, 90% have already connected AI tools to internal systems. The findings suggest that AI is increasingly being embedded within firms’ operational infrastructure rather than being used as a standalone productivity tool.
- Agentic AI Is Already Emerging
Agentic AI is also moving into production. The Survey found that 30% of respondents have already deployed agentic AI in production environments, while a further 25% are conducting controlled pilots. Use cases include information retrieval, drafting, software development and workflow automation. Although more autonomous systems are emerging, firms generally continue to maintain varying degrees of human review and oversight over AI-driven activities.
- Connectivity Expands the Control Perimeter
One of the Survey’s most significant observations is the extent to which AI tools are being integrated with internal systems. Connections to messaging platforms, document repositories, customer data systems and trading-related systems significantly expand AI’s operational footprint. As AI becomes more deeply connected to business processes, the control perimeter expands with it, bringing access rights, confidentiality, data quality, system permissions and downstream decision-making into the governance framework.
- Governance Is Emerging as the Key Scaling Challenge
Taken together, the Survey findings suggest that the challenge is increasingly shifting from demonstrating AI capability to scaling AI within robust control frameworks. Firms reported difficulties in scaling control frameworks, maintaining accountability, validating outputs and moving pilot projects into resilient production environments.
As AI applications become more interconnected and autonomous, regulators are likely to focus more closely on whether governance arrangements evolve at the same pace as technological deployment.
The SFC’s Supervisory Priorities
Drawing on the Survey findings, the SFC identified four supervisory priorities:
- investor protection;
- cybersecurity and operational resilience;
- effective AI governance with embedded human accountability; and
- industry engagement and collaboration.
Importantly, human accountability does not necessarily mean human approval of every AI-generated output or action. Rather, controls should be proportionate to the authority and potential impact of the relevant AI application.
Takeaways
The Survey suggests that supervisory focus is shifting from AI adoption to AI governance, particularly as AI becomes more deeply integrated into firms’ operations, systems and decision-making processes.
Licensed corporations should therefore consider whether they have:
- a complete inventory of AI applications, including embedded and third-party tools;
- risk-based approval and oversight arrangements proportionate to each application’s autonomy and potential impact;
- appropriate controls over data access, confidentiality and system integration;
- clear human accountability for AI-supported decisions and workflows; and
- testing, monitoring and escalation procedures that remain effective as AI use expands.
For financial institutions, the emerging regulatory question is therefore no longer simply “Can AI perform the task?”, but “Can the firm demonstrate that the task remains properly governed?”
Please contact our Partner Mr. Rodney Teoh for any enquiries or further information.
This news update is for information purposes only. Its content does not constitute legal advice and should not be treated as such. Stevenson, Wong & Co. will not be liable to you in respect of any special, indirect or consequential loss or damage arising from or in connection with any decision made, action or inaction taken in reliance on the information set out herein.
