Agentic AI surges in HK finance as 95% expect faster adoption

Agentic AI is gaining real traction in Hong Kong's financial sector, with 95% of firms expecting their AI use to accelerate over the next 12 months, and its reach already spans front-, middle- and back-office activities, a Securities and Futures Commission (SFC) survey finds.

The "Large firm AI readiness survey" highlights broad use of artificial intelligence (AI) across business and control functions, alongside challenges to scaling its use responsibly. It was conducted between July and August 2026, covering 20 large financial firms with 78 affiliated licensed corporations in Hong Kong.

Their responses indicated that AI use is extending across business and control functions, becoming more closely integrated with internal systems and, in some applications, operating with greater autonomy. Notably, agentic AI is gaining traction with major use cases covering information retrieval, drafting, software development, operational workflows and some trading-related activities. 30% of the surveyed firms already use agentic AI in their business operations and 25% are piloting it, with varying degrees of autonomy and human involvement.

On the front line, AI is most widely deployed in investment research and analytics (80%), followed by the preparation of client communications or documentation and client service and communication management (65% each). Adoption is less widespread in marketing, advertising and promotion (45%), trading and execution and customer analytics (30% each), product offering and distribution (20%), and investment advice and portfolio management (15%). This pattern indicates a broader reported footprint in research, information preparation and service support than in investment advice, portfolio management or execution.

AI adoption is gaining momentum, with 95% of surveyed firms expecting their AI use to accelerate over the next 12 months. Its reach already spans front-, middle- and back-office activities, and integration is deepening, with 90% having connected AI tools to internal systems. Firms are also leveraging AI for cyber defence: 90% are using or evaluating frontier AI models for this purpose, reporting benefits in faster analysis and broader detection coverage.

However, practical barriers to broader adoption remain: technology integration, the transition from pilots to production, costs and uncertain business value, data constraints, and skills shortages. Firms also struggle to extend governance and controls across a growing number of applications while maintaining clear accountability and human oversight.

Technology integration, infrastructure and controlled scaling is the most cited barrier, flagged by 70% of firms. Cost, resources and business-case development follows at 60%, while data quality, availability, privacy and security; regulatory, compliance and cross-jurisdiction alignment; and skills, workforce adoption and change management are each cited by 55%.

The survey findings will inform the SFC's supervisory approach to AI, with four priorities: protecting investors as AI reshapes their interactions with firms; strengthening cybersecurity and operational resilience as AI introduces new threats and dependencies; ensuring effective AI governance by rethinking how human accountability and oversight are embedded in increasingly autonomous workflows; and building knowledge and capability through engagement and collaboration.

The SFC has reminded firms to comply with applicable legal and regulatory requirements to support the responsible adoption of AI in Hong Kong's capital markets. It will focus on preserving human accountability throughout AI-enabled business processes, with clear objectives, defined limits on access and authorised actions, and effective monitoring and intervention. Controls should be proportionate to an application's authority and potential impact, rather than requiring manual approval for every output or action.

Eric Yip, the SFC’s executive director of Intermediaries, said: “Our supervisory approach is safety-first, human-anchored and collaboration-led. As AI becomes more autonomous, we must put humans even more firmly in the driving seat. Scaling AI takes more than powerful models and promising use cases. It takes reliable infrastructure, sound data, the right expertise and governance that keeps pace with the technology.“

Related articles:

Streaming fatigue meets AI anxiety in Deloitte's media report
AI marketing platform Profound plants APAC flag in SG amid AI search boom
HKTV boss Ricky Wong on the power of AI and inevitable job losses