What is one specific practice or element of data governance that is essential for organizations to safely leverage AI with their enterprise data, and why?
Most answers emphasize the need for clear data classification, unified governance, and consistent definitions to ensure AI uses enterprise data safely and reliably, though some focus on quality, ownership, or guardrails.
10 answers
Answers


Shayde Christian
Chief Data & Analytics Officer · Cloudera
Consistent Naming Conventions
Shayde Christian says his team provided naming conventions, renaming tables and columns so AI understands business meaning like annual recurring revenue, because AI needs more help to grasp how data is described and governed.
The answer starts at 27:06.
Chain of Thought | AI Agents, Infrastructure & Engineering · Listen on Apple Podcasts
Dana Simberkoff
Chief Risk, Privacy and Information Security Officer · AvePoint
Policy and Governance Process
Dana Simberkoff says organizations need a policy in place followed by a governance process for approving future AI use cases, because no two implementations fit the same stack.
AvePoint · Watch on YouTube
Jim Rowan
Principal (Equity Partner) · Deloitte
Master data management
Jim Rowan says organizations are investing in master data management layers needed to drive AI, part of building the data foundation and controls that ground AI in consistent definitions of customers and products.
The answer starts at 6:29.
DataFramed · Listen on Apple Podcasts

Casey Ellis
Principal · Tall Poppy Group
Enforcing data classification
Casey Ellis says enforcing data classification is harder than labeling it, since deciding who gets access and managing exceptions is difficult, and employees may not know how data should be classified without a subject matter expert.
Security Weekly - A CRA Resource · Watch on YouTube

Anahita Tafvizi
Chief Data & AI Officer · Snowflake
Unified governance
Anahita Tafvizi says organizations need unified governance and unified metric definitions so AI gives trusted answers, warning that without it AI can scale chaos across the company.
New York Stock Exchange · Watch on YouTube
Stefan Zima
Head of Group Regulatory Compliance / Dep. (BWG) Regulatory Compliance Officet · Raiffeisen Bank International Ag
Guardrails
Stefan Zima says organizations must govern AI hard and guardrail their use case and AI tooling worlds, citing an OpenAI example of a model finding an easier route to pass a test.
The answer starts at 4:26.
The AI in Business Podcast · Listen on Apple Podcasts
Mike Ferguson
Owner, CEO and Head of Research · Intelligent Business Strategies
Data classification
Mike Ferguson says companies must now classify data not only for sensitivity rules like GDPR but also to find data and collections that satisfy an AI agent's relevance needs at a given time.
The answer starts at 4:40.
It's About Data · Listen on Apple Podcasts

Brennan Baybeck
SVP & CISO - Customer Success Services · Oracle
Data classification
Brennan Baybeck says data classification has never been more important as part of strong governance, calling it a foundational security principle and a necessary change, with strict classification and policy enforcement needed to filter data moving in and out of AI models.
SiliconANGLE theCUBE · Watch on YouTube

Sam Pierson
Chief Technology Officer · Qlik
High‑quality data
Sam Pierson says having a data foundation that is quantifiably reliable and high quality matters, and that adding context on top of the data brings the quality and accuracy of LLM answers to the next level.
Data Faces Podcast · Data Faces Podcast · Listen at 8:12

Manish Sood
Chief Executive Officer, Founder & Chairman · Reltio
Data ownership
Manish Sood says an organization's data is its unique asset and that ownership is theirs, framing how they make it available for use as the way to innovate and inform AI outcomes.
Cloud Wars Live with Bob Evans · Cloud Wars Live with Bob Evans · Listen at 14:14

