The rush by financial institutions to implement new artificial intelligence-powered applications is contributing to a broad rewriting of the rule book on data management, particularly how that data is governed.
New frameworks are having to be created to govern data as AI tools transform not only the way data can be generated and ingested but also the way it is distributed. As a result, privacy and security safeguards are having to be redrawn.
A key theme within the discussion of modern data governance is how AI is challenging the rigid top-down approach that has characterised frameworks in the past; no longer is the organisation the creator and commander of data – AI is enabling users to create enterprise-useful information. That, in turn, has introduced new risks; how can the organisation track and manage the creation of this new body of data?
Deep Dive
In A-Team Group’s upcoming webinar, “Hearing from the Experts: AI Governance Best Practices”, experts drawn from capital markets and their suppliers will take a deep dive look at this fascinating and rapidly evolving topic.
Nicole Hansen, Compliance and Conduct Lead at NatWest Group; Joanne Biggadike, Head of Data Governance at Schroders; and Gaston Hummel, EMEA Lead Strategic Services at Precisely will discuss not only how data governance is being reshaped by AI but also examine some of the best approaches to solving the challenges it has created.
Ahead of the webinar, Data Management Insight spoke to two of the panellists about why new governance structures are needed and how they can benefit AI applications.
Governance is critical for AI
Joanne Biggadike, Schroders: Generative AI relies on high-quality, reliable data to function effectively. While governance in the context of AI is not a new concept, what is relatively new is the scale and accessibility of AI. This increased accessibility heightens the importance of robust governance frameworks.
Gaston Hummel, Precisely: AI governance efficiently democratises a common understanding of critical AI assets and their relationships. For successful and responsible AI, asset types to be governed and well understood across your organisation include AI use cases, AI models, metrics (business impact, quality, compliance), including measurement details, current and target values, data and its sovereignty plus lineage, regulations, policies, and more.
AI governance requires innovative new approaches
Jo Biggadike: It is important to recognise AI governance is distinct from data governance. AI governance should address not only the purpose and justification for AI usage, but also whether it should be used at all. Effective AI governance requires broader stakeholder involvement and oversight. Strong data governance provides the essential foundation for this by ensuring that data can be interrogated, reviewed, and validated.
The benefits of getting governance right are many-fold
Gaston Hummel: AI governance empowers your whole organisation to understand and confidently trust its AI’s use cases, AI’s output, AI’s limitations and a clear understanding of AI’s potential and actual impact across your organisation, both positive and negative. Proper AI governance empowers and democratises responsible, transparent, efficient and measurable AI usage that is aligned with corporate objectives and provides a framework for continuous improvement and reporting. Furthermore, it clearly assigns responsibilities and operationalises structured operating models for transparent AI use case collaboration and rapid iteration. AI governance provides the framework and operating model to accelerate responsible AI usage to help achieve organisations’ goals.
- A-Team Group’s “Hearing from the Experts: AI Governance Best Practices” will be moderated by Data Management Insight editor Mark McCord and will take place on September 9 at 10:00am ET/3:00pm London/4:00pm CET. Click here to register your attendance.
Subscribe to our newsletter
