Talk

From MCP to Agentic AI: Can We Trust Autonomous Agents to Uphold Ethics?

3 min read
Poster for the 2026 IEEE Computer Society Malaysia Chapter Webinar Series session, From MCP to Agentic AI: Can We Trust Autonomous Agents to Uphold Ethics?, 27 October 2026

The IEEE Computer Society Malaysia Chapter, together with the Centre of Research for Cyber Security & Network (CSNET), Universiti Malaya, continued the 2026 IEEE Computer Society Malaysia Chapter Webinar Series & Professional Workshops with a session on the ethics of autonomous AI agents.

Session Details

Watch the Recording

The full session, including the Q&A, is available on YouTube: https://youtu.be/1hTW9npbaJs

Synopsis

As AI agents gain the ability to connect to external tools and data through protocols such as the Model Context Protocol (MCP), they are moving from answering questions to acting on our behalf. This shift raises a practical question: can autonomous agents be trusted to uphold ethics when they plan, decide and act with limited human oversight?

This session traces the path from MCP to agentic AI and examines the ethical issues that arise along the way, including accountability, transparency, privacy, bias and safe human oversight. Designed for professionals, researchers and students, it offers clear explanations and practical considerations for building and using trustworthy agents.

Speaker

Professor Dr Siti Hafizah Ab Hamid, Deputy Dean, Universiti Malaya.

Moderator

Ts. Dr. Muhammad Faiz Bin Mohd Zaki, Director, Data and Information Management Centre, Universiti Malaya.

Key Learning Objectives

  1. Understand the Path from MCP to Agentic AI. Explore how protocols such as MCP connect models to tools and data, and how this enables autonomous agents.
  2. Recognise Ethical Risks of Autonomous Agents. Identify issues of accountability, transparency, privacy and bias when agents act on their own.
  3. Assess Trust in Agent Decisions. Learn what it means to trust an agent, and how to evaluate whether its actions are reliable and aligned with human values.
  4. Apply Safeguards and Human Oversight. Discover practical techniques, such as permissions, audit trails and approval steps, that keep agents within clear boundaries.
  5. Prepare for Responsible Adoption. Identify guidelines and good practices for using agentic AI responsibly in research, work and daily tasks.

← Back to all news