AI governance is no longer optional for Hong Kong SMEs. Professional firms—accountants, lawyers, TCSPs, and financial boutiques—are already using AI for client communication, research, and document handling. The challenge is not adoption, but control.

Regulators like the PCPD are clear: firms should implement governance, risk checks, monitoring, and clear usage rules before AI becomes embedded in daily work. At the same time, market pressure is rising. Over 50% of SMEs now prefer candidates with AI skills, and misuse of AI tools is increasingly linked to data leakage risks.

For SMEs, a practical approach starts with three questions:

• Where is client data stored?

• Which AI tools are approved?

• How is work tracked and verified?

A simple operating model can address this. HubSpot acts as the system of record for client relationships, tracking interactions, onboarding stages, and decisions. Perplexity supports research and knowledge work, helping teams produce sourced, structured outputs without relying on scattered tools.

Take client onboarding as an example. Instead of fragmented emails and spreadsheets, HubSpot structures the workflow—from enquiry to approval—with clear visibility and audit trails. Perplexity can assist with regulatory research, summarising public information, and drafting internal checklists, while keeping outputs traceable.

An effective SME AI policy does not need to be long. It should define:

• Approved tools

• Prohibited data inputs

• Verification requirements

• Record-keeping standards

At its simplest: do not input sensitive data into unapproved tools, verify AI outputs, and log client-impacting work in your CRM.

The firms that benefit most from AI will not be those using the most tools, but those with clear workflows and governance. HubSpot provides structure, Perplexity enables informed work, and together they create a controlled, scalable way to adopt AI.