Enterprise AI systems can run on flawless data pipelines and still produce answers nobody trusts. Beth Hiatt, SVP of Product Management, Data Governance and Responsible AI at LPL Financial, continues her conversation with Jessie Smith, Chief Product Officer at Ataccama, for the second installment of a three-part series on where that trust actually breaks down.

Hiatt shares her perspective that even clean, well-governed data does not guarantee a trustworthy AI answer. Data quality, traceability and definitions can all be in order, and a model can still miss the mark because nobody built in the business context for what its output needs to deliver.

That same gap is reshaping data stewardship. Hiatt points to a four-part CDO Magazine series she helped write, “The Digital Data Steward,” built on a term she credits to her longtime mentor, Maria Villar. The series argues that stewardship no longer requires a human checking every record by hand.

• Context over quality: Even the highest-quality data can fail an AI use case without the business context behind it.
• Agents doing the steward’s work: Metadata hydration, data quality checks and classification are increasingly agent-led rather than manual.
• A CDO’s view of risk and opportunity: From a Chief Data Officer’s (CDO) perspective, Hiatt says agents can flag where data risk is highest and where the data points to new business opportunities.

Did you miss the previous episode?

On the move? Take this conversation with you. If you don’t have time to finish the video, you can stream the complete, uninterrupted audio edition of this episode on the go.