Data trust issues block AI deployment in Hong Kong

Governance, data trust and talent gaps slow rollout.

Hong Kong businesses are accelerating adoption of artificial intelligence (AI) agents, but scaling beyond pilot projects remains constrained by weak data foundations, talent shortages, and governance challenges.

Recent industry findings show that whilst 85% of firms are expanding AI use, only a small share of projects reach full production, underscoring a widening gap between experimentation and real-world deployment. Kunal Taneja of Databricks said 2026 marks a turning point. “In 2026, it marks a transition from AI experimentation to embedded production grade AI that delivers ROI,” he said, noting a shift towards systems that are “reliable, governed, and business aware.”

Adoption of more advanced systems is rising rapidly. “Customers grew adoption of multi-agent deployment by more than 300% coming into 2026,” Taneja said, with nearly 90% of use cases focused on real-time applications. These systems are enabling faster decision-making, with one deployment reducing market research timelines from weeks to minutes.

However, scaling remains uneven. “If you can't trust the data, it is hard to trust what you're going to get back from AI systems,” Taneja said, highlighting how legacy infrastructure and siloed systems continue to undermine adoption. Talent constraints further complicate progress, with “more than 25% of organisations” struggling to hire workers with AI and data science expertise.

Governance is emerging as a decisive factor. Hong Kong’s regulatory environment emphasises “explainability, accountability and data governance,” but compliance requirements can slow deployment. At the same time, firms that invest in governance frameworks are seeing stronger outcomes, with some achieving significantly higher rates of production deployment.

Professor Zhang Jiheng of Hong Kong University of Science and Technology said broader structural barriers persist. “The general public do not know AI as much,” he said, pointing to literacy gaps alongside integration challenges, security risks, and the disconnect between digital systems and physical operations.

Despite these challenges, certain sectors are advancing faster. Financial services firms, supported by structured data and established workflows, are leading adoption. One insurer recorded a doubling of customer engagement and lead generation through AI-driven analytics, whilst telecom and logistics firms are exploring applications in fraud detection and supply chain coordination.

The trajectory is clear: AI adoption is accelerating, but scaling depends on governance, data trust, and talent readiness. For businesses, the next phase will be defined less by experimentation and more by their ability to operationalise AI within secure, integrated, and well-governed systems.