Background
AI is moving from a tool that generates content to an agent that acts. Agentic AI systems can plan and execute multi-step tasks such as negotiating, purchasing, or processing transactions with limited human input, across company systems.
This shifts who a business serves (customers’ AI agents as well as customers), how work gets done, and where risk sits. It also aligns with Hong Kong’s direction: the Five-Year Plan for Economic and Social Development (2026-2030) positions AI as a core industry and promotes “AI+” integration across sectors, alongside AI governance, data security, and the commercialisation of technology. The question for businesses is no longer whether to adopt AI, but how to rethink value creation when AI can act on their behalf, on their customers’ behalf, or against their interests.
What We Mean by “Agentic AI”
Systems that can pursue a goal by planning steps, using tools or systems, and taking actions with limited human involvement. A chatbot that answers questions or a copilot that drafts text does not qualify on its own. Proposals must involve AI taking meaningful autonomous action.
The Challenge
Choose a company whose business could be materially reshaped by AI agents. Identify where agentic AI creates the greatest opportunity or threat, and propose a strategy the company can realistically execute.
Submission Details
1. Choose an existing listed company in Hong Kong or Greater China. Explain why it is specifically exposed to agentic AI.
2. Choose one primary track. You may touch on the others where they interact, but your strategy should be anchored in one.
- Agentic Commerce. Customers increasingly delegate search, comparison, and purchasing to AI agents. How does a company stay visible, trusted, and profitable when the “customer” is an algorithm, and how can it deploy its own agents to sell, personalise, and serve better?
- Agentic Cybersecurity. Autonomous agents create new attack surfaces, fraud vectors, and accountability gaps. How should a company manage the business consequences (financial loss, liability, regulatory exposure, customer trust) of agents acting inside and outside its walls?
- Agentic Workforce. When agents can execute work, what should be done by humans, by agents, or by both together? How should roles, decision rights, accountability, and skills be redesigned?
3. Diagnose the threat or opportunity. Analyse the company’s business model, operations, and capabilities. Pinpoint the clearly defined threat, or opportunity created by agentic AI, supported by evidence.
4. Propose and justify the strategy. Show what the company should do, how it should do it, and why it creates value.