From data quality to incident response—keep every pipeline https://clomidxx.com/report-massachusetts-general-hospital-targeting-various-blockchain-use-cases/ trustworthy at scale. Learn how to advance ethical and compliant AI practices through a unified set of generative AI governance capabilities. A strong data foundation is critical for the success of AI implementations. IBM watsonx.ai is an enterprise-ready studio, bringing together traditional machine learning (ML) and new generative AI capabilities powered by foundation models. The data governance solution powered by IBM Knowledge Catalog offers several capabilities to help facilitate advanced data discovery, automated data quality and data protection. While grounding the conversation in today’s newest trend, agentic AI, this AI Academy episode explores the tug-of-war that risk and assurance leaders experience between governance and security.
Their initial task is to improve data quality management for a single, high-impact use case. The first step isn’t to boil the ocean but to start with a targeted approach focused on business value. Our organization’s data is inconsistent and siloed. Data governance focuses on managing data as an asset throughout its lifecycle—establishing policies for data quality, access, and security. Are drift alerts and ethical red flags routed to accountable owners within SLA?
Business value comes from better decisions grounded in evidence, trusted information and clear accountability. Mr. Marco established his authority in enterprise data management long before the current AI era. The distinction matters because current automated governance technology cannot operate without human intervention, particularly for policy implementation, so full autonomy is not a credible claim. Human on the loop means agents handle the drafting, classification, and documentation work while human stewards approve what ships, set the thresholds, and intervene on exceptions. Atlan’s MCP server and conversational AI serve context to humans and agents at inference under identical persona entitlements.
Measuring Success: KPIs That Prove Business Value
Because of this, when we look to manage and govern the deployment of AI models, we must first focus on governing the data that the AI models are trained on. For the most part, artificial intelligence models are fairly simple, they take in data and then learn patterns from this data to generate an output. Senior Product Manager – Data privacy and regulatory compliance
- Data governance for AI ensures responsible, secure, and compliant data management throughout the AI lifecycle, from training to deployment.
- AI governance is the set of rules, institutions, and internal controls that determine how AI systems are built, deployed, and held accountable.
- Organizations should review their AI governance framework at least quarterly, since AI-generated content volume, agent deployments, and unsanctioned tools usage all accumulate continuously between review cycles.
- Who is ultimately responsible for data and AI governance in an organization?
- For nearly three decades, Mr. Marco has worked at the intersection of enterprise data and executive accountability.
Human on the loop means agents do the drafting, classification, and documentation work while stewards approve what ships and intervene on exceptions. When a business rule changes, you change the rule in the semantic layer and every agent reading it adapts. Certification status needs to be as machine-readable as the PII flag, because “is this the approved revenue metric” is the question agents get wrong most often. Gartner has predicted that over 40% of agentic AI projects will be canceled by the end of 2027, due to escalating costs, unclear business value or inadequate risk controls. AI systems present unique governance challenges that traditional data management approaches cannot fully address. Data governance for AI ensures responsible, secure, and compliant data management throughout the AI lifecycle, from training to deployment.
Learn How to Build AI Trust
Running one lineage graph, one policy layer, and one audit trail across data, models, and context artefacts is what makes an audit answerable rather than a reconstruction project. GDPR obligations connect directly to AI obligations, so an output produced by a model reading a customer table sits under both regimes at once. Unlike predictable traditional systems, AI outputs can be chaotic and difficult to test comprehensively, necessitating continuous monitoring and specialized controls. Store it in open formats you can query from your own warehouse, so the next model, the next agent framework, and the next vendor decision cost you a connector rather than a rebuild. To scale data governance for AI, govern one lifecycle, and not three. Checks run in-warehouse, so there is no new compute and no data leaves the perimeter, and AI drafts first-version rules so coverage reaches assets nobody would have written tests for by hand.
- EU AI Act Article 10 establishes data-governance duties for covered high-risk systems.
- Exceptions route to stewards, drift triggers review, and the graph updates rather than the model getting retrained.
- Classification is what turns a generic access policy into an enforceable one.
- The EU AI Act creates enforceable duties for covered high-risk systems, while the NIST AI Risk Management Framework remains voluntary guidance that many organizations use to structure controls.
Existing capabilities, such as metadata management, data quality processes, and stewardship roles and activities, provide a foundation that can be extended into AI governance. AI governance is no longer aspirational; it is an operational requirement for organizations seeking to scale AI responsibly and sustainably. Many organizations expect data leaders to address this concern without a defined playbook. Today, he applies that https://ulstergrandprix.net/category/news/page/18/ foundation to Decision Intelligence, focusing on how organizations make high-consequence decisions in an environment increasingly shaped by AI. He helps leaders align accountability and decision rights, ensure the information behind key decisions can be trusted, and make decisions that deliver business value and withstand scrutiny. David Marco advises boards and C-suite leaders on Decision Intelligence, helping them make better business decisions by aligning AI, trusted information, governance and executive accountability.
- Because of this, when we look to manage and govern the deployment of AI models, we must first focus on governing the data that the AI models are trained on.
- AI governance is no longer aspirational; it is an operational requirement for organizations seeking to scale AI responsibly and sustainably.
- Atlan’s Context Engineering Studio derives test cases from downstream lineage, so the questions an agent is graded on reflect what people actually ask of that data.
- Continuous human and system monitoring can ensure AI systems perform as expected and remain aligned with policy requirements.
Step 5: Develop Policies, Standards, and Controls
A comprehensive inventory of AI systems is the optimal starting point for governance development, to support scope and program design. Clarity at this stage prevents ambiguity and ensures that governance efforts are supported at the executive level and are focused and actionable. Organizations must determine which AI systems, business units, and use cases fall within the initial effort, recognizing that scope can expand over time.