CPTEnglishIntermediate· Artificial Intelligence, AI Governance

From Chatbots to Agents: The Phase of AI

Mastering the shift from conversational chatbots to goal-oriented autonomous AI agents

Legal & Compliance 1h 30 lessons

What you'll learn

Differentiate between reactive chatbots and proactive, goal-directed AI agents
Identify the core components of agent architecture, including planning, memory, and tool use
Analyse the shift from content generation to operational execution and its impact on risk
Apply governance frameworks to manage permissions, accountability, and human oversight
Explain the 'agent loop' of planning, executing, observing, and revising tasks
Evaluate potential use cases for automating enterprise workflows and software development
Design strategies for maintaining security and privacy within agentic memory systems

About this course

This CPT course explores the evolution of artificial intelligence from conversational chatbots to autonomous goal-directed agents. For Hong Kong SFC-licensed professionals, understanding this transition is critical as financial institutions and family offices increasingly explore agentic workflows for trade execution, compliance monitoring, and administrative efficiency. While chatbots were largely limited to information retrieval and content generation, AI agents operate in systems of action, necessitating a robust understanding of how these tools can be integrated into highly regulated environments while maintaining operational stability and compliance with internal control requirements.

Who should attend

Compliance Officers
Risk Managers
Licensed Representatives
FinTech Operational Leads
Family Office Investment Managers
Data Governance Professionals
IT Audit Specialists

Key topics covered

Foundations of Agentic AI

Distinguishing between conversational interaction and goal-directed operational performance
The agent loop: planning, executing, observing, and revising tasks
Core components including tool integration and long-term memory systems
Structuring automated workflows within the enterprise environment

Governance and Risk Management

Mitigating goal drift and hallucination in decision-making workflows
Defending agentic systems against prompt injection and security vulnerabilities
Implementing resource limits and stop conditions for safe operation
Managing automation complacency and overtrust in human-agent collaboration

Operational Oversight and Compliance

Designing robust permission architectures for secure tool access
Implementing human-in-the-loop models for high-stakes decisions
Ensuring accountability through comprehensive agent logging and audit trails
Navigating policy frameworks for governing delegated actions in financial services

How this course satisfies CPT

This course covers Artificial Intelligence, AI Governance topics recognised by the SFC as eligible for CPT. Upon completion, you'll earn approximately 1 CPT hours and receive a certificate of completion for your training records. The course includes quizzes to verify your understanding, satisfying the SFC's requirement that online learning includes an independent assessment and maintains a reliable audit trail.

Learn more about CPT requirements

Curriculum · 30 lessons

1:07:31.0 of content

1.
Transitioning from Generative AI Chatbots to Autonomous AI Agents
FREE
2:02.3
2.
Understanding the Fundamental Shift from Language Output to Goal-Directed Action
FREE
2:01.9
3.
Governance Strategies for AI Agents Operating in Systems of Action
FREE
2:19.8
4.
Managing Risk and Accountability in Autonomous AI Agent Workflows
2:19.8
5.
Distinguishing Between Chatbot Interaction and Agentic Operational Performance
2:11.8
6.
Quiz: Transitioning from Generative AI Chatbots to Autonomous AI Agents – Distinguishing Between Chatbot Interaction and Agentic Operational Performance
QUIZ
5 questions · requires lessons
7.
Defining Key Components of AI Agents: Goals, Tools, and Oversight
2:08.8
8.
The AI Agent Loop: Planning, Executing, Observing, and Revising Tasks
2:08.2
9.
Integrating Tools to Transform AI from Conversationalist to Operational Actor
2:10.7
10.
Designing Secure and Contextual Memory Systems for AI Agents
2:17.7
11.
Enhancing Agentic Performance Through Structured Planning and Reasoning Models
2:17.7
12.
Quiz: Defining Key Components of AI Agents: Goals, Tools, and Oversight – Enhancing Agentic Performance Through Structured Planning and Reasoning Models
QUIZ
5 questions · requires lessons
13.
Automating Enterprise Workflows with Practical AI Agent Implementations
2:23.3
14.
Optimizing Team Coordination Using AI Agents as Digital Co-workers
2:11.9
15.
Leveraging AI Agents to Augment Software Development and Code Quality
2:19.9
16.
Accelerating Scientific and Technical Discovery Using Research-Focused AI Agents
2:24.6
17.
Improving Clinical and Administrative Efficiency with Healthcare AI Agents
2:26.6
18.
Quiz: Automating Enterprise Workflows with Practical AI Agent Implementations – Improving Clinical and Administrative Efficiency with Healthcare AI Agents
QUIZ
5 questions · requires lessons
19.
Personalized AI Assistants for Managing Complex Daily Life Tasks
2:09.6
20.
Navigating the Safety Challenges of Embodied AI Agents in Physical Spaces
2:15.4
21.
Addressing Complexities in Building Agents for Unstructured Real-World Tasks
2:16.0
22.
Mitigating Goal Drift to Keep AI Agents Aligned with Intent
2:13.1
23.
Implementing Resource Limits and Stop Conditions for Reliable AI Operation
2:13.8
24.
Quiz: Personalized AI Assistants for Managing Complex Daily Life Tasks – Implementing Resource Limits and Stop Conditions for Reliable AI Operation
QUIZ
5 questions · requires lessons
25.
Managing the Risks of Hallucination in Agentic Decision-Making Workflows
2:10.7
26.
Defending AI Agents Against Prompt Injection and Security Vulnerabilities
2:13.4
27.
Addressing Bias and Fairness in Scalable AI Agent Decision Systems
2:26.2
28.
Strategies to Prevent Automation Complacency and Overtrust in AI Agents
2:13.2
29.
Designing Robust Permission Architectures for Secure Agentic Tool Access
2:11.4
30.
Quiz: Managing the Risks of Hallucination in Agentic Decision-Making Workflows – Designing Robust Permission Architectures for Secure Agentic Tool Access
QUIZ
5 questions · requires lessons
31.
Implementing Effective Human Oversight Models for High-Stakes Agentic Decisions
2:07.3
32.
Ensuring Auditability and Accountability Through Comprehensive AI Agent Logging
2:13.8
33.
The Role of Technical Standards in Safe Agent Interoperability
2:25.5
34.
Policy Frameworks for Governing Delegated Action in the Agent Era
2:20.7
35.
Building Trustworthy AI Agents for the Next Phase of Operation
2:16.0
36.
Quiz: Implementing Effective Human Oversight Models for High-Stakes Agentic Decisions – Building Trustworthy AI Agents for the Next Phase of Operation
QUIZ
5 questions · requires lessons

Requirements

Fundamental knowledge of generative AI and large language models
Interest in enterprise workflow automation and AI operations
Basic understanding of software development lifecycle or business process management
Awareness of data security and organizational compliance principles

Certificate of Completion

Complete this course to earn a verifiable Certificate of Completion for your CPT training record. Here's how:

1Watch all the lessons — work through every lesson in the curriculum. Each completed lesson brings your progress closer to 100%.
2Pass all the quizzes — score 70% or higher in each quiz within the course. Quizzes are part of the curriculum and must be passed to finish.
3Download your certificate — once you reach 100% progress, download a personalised Certificate of Completion showing your name, the CPT hours earned, and your quiz score.

Sample certificate: