Feeding the AI Machine: Training Data Under Regulatory Scrutiny in Asia

As artificial intelligence model performance races ahead, regulators increasingly focus less on algorithms and more on one question: Where did the training data come from, and can you justify its reuse? This panel examines how that question arises in audits and investigations across China, Korea, Thailand and Vietnam. Expectations are often principles-based, requiring clear narratives on provenance and purpose f it. Through practical scenarios, speakers will show how companies and institutions can responsibly enable training-data use, clarifying when and how legacy repositories, scraped sources and third-party datasets can be used under regulatory scrutiny. Rather than surveying every rule, this session distills cross-jurisdiction themes regulators repeatedly test: provenance, purpose-fit decisions and internal accountability. It will also cover who must answer, and with what documentation.

What you will learn:

• Where training-data obligations sit in China, Korea, Thailand and Vietnam statutes, and what they practically mean.

• The first questions regulators ask, the triggers that make reuse contested, and com-mon failure points in audits and investigations.

• How to document and govern training-data decisions in a way regulators expect to see across jurisdictions.