On March 19, 2026, the Centre for Teaching and Learning (CTL) successfully hosted the tenth session of the "Beyond Boundaries" Teaching and Learning Lecture Series. More than 80 faculty members and students attended this engaging session on agentic AI. Dr. Wentao CHENG, Associate Professor in the Department of Computer Science at BNBU, delivered the lecture titled OpenClaw: Introduction, Deployment, and Security Considerations for Modern Automation. He guided the audience through the practical steps of building a personal AI assistant and emphasized the importance of data sovereignty and secure deployment.

Dr. CHENG offered a unique perspective shaped by both industry and academia. Currently an Associate Professor at BNBU, he previously held the same position at Nankai University. Before entering academia, he worked as a senior algorithm engineer and product manager at Alibaba Group, where he developed a multi-sensor perception system for autonomous vehicles. This combination of hands-on engineering practice and academic research gave him deep insights into the real-world deployment of AI systems, which he shared throughout the lecture.

Dr. Wentao CHENG speaking at the lecture

In his lecture, Dr. CHENG challenged the common view of AI as merely a question-and-answer tool, arguing that we are entering an era of truly autonomous digital assistants. He demonstrated how frameworks such as OpenClaw enable these agents to complete real-world tasks independently, from curating personalized news to managing complex research workflows. Yet with greater autonomy comes greater responsibility. Dr. CHENG warned that as agents gain more independence, users must remain vigilant about security, treat external information with caution, and always retain final decision-making authority.

Participants listened to the lecture attentively

To translate this vision into practice, Dr. CHENG introduced a practical approach based on local deployment and user ownership. He stressed that running these agents on personal hardware allows individuals to regain control over their data. He also highlighted the value of a human-in-the-loop model, in which agents handle routine workloads while humans provide judgment and final approval. He concluded by encouraging the audience to build with caution, automate with purpose, and remember that owning your agent means owning your digital future.

Q&A session

The Q&A session saw active participation, with attendees raising thoughtful questions about deployment practices and security measures. Drawing on his professional experience, Dr. CHENG provided practical insights and encouraged the audience to approach agentic AI with both enthusiasm and caution.

This lecture successfully demystified agentic AI, shifting the focus from passive chatbots to proactive digital agents that users can fully own. By emphasizing local deployment and data sovereignty, it equipped faculty and students with both practical skills and a security-focused mindset for purposeful automation. Going forward, CTL will continue to deliver thought-provoking, transdisciplinary dialogues through the "Beyond Boundaries" Lecture Series, steadily advancing teaching quality and the holistic development of the academic community.