In June, I had the privilege of speaking at the DGIQ West conference, a key event for professionals focused on AI governance and data product management. As Kenway Consulting’s Data Governance Service Lead, and part of an organization named a “Pacesetter” in ALM Compass Research’s 2024 Data Governance Report, it was a great opportunity to discuss ideas with so many like-minded, passionate colleagues. This year’s conference highlighted breakthroughs in four crucial areas that are poised to transform how organizations utilize AI:
Key Takeaway: Applying product management principles to AI as a data product is essential for maximizing its strategic value and addressing complex challenges effectively
We've often discussed product management in the context of tangible goods and services, but applying this discipline to data, particularly AI, is a game-changer. Treating AI as a data product means recognizing it evolves over time. As user personas and use cases for AI evolve, robust product management practice becomes essential. This approach can significantly enhance the application of AI, ensuring it addresses strategic and complex challenges rather than merely being a tool to avoid the fear of missing out (FOMO).
Key Takeaway: Effective AI governance is critical and urgently needed. By applying data governance principles, organizations can establish a robust framework for ethical and strategic AI deployment.
A recurring theme at the DGIQ West conference was the urgent need for AI governance. Despite its critical role, our current efforts in AI governance are insufficient. Data governance, however, is well-positioned to fill this gap. The core principles of data governance provide a solid foundation for AI governance, allowing us to make informed decisions and establish guardrails for responsible AI use.
One example of how data governance can evolve into AI governance is through guiding principles. Data governance has long focused on the responsible use of data. AI governance must take this a step further, addressing the ethical implications of AI use. As AI becomes more pervasive, ensuring its ethical application is a natural progression from traditional data governance.
Another critical aspect is the type of data involved. Data governance has traditionally excelled with structured data, whereas AI predominantly deals with semi-structured and unstructured data. Despite these differences, the foundational principles of data governance still apply. Adapting these principles to the unique challenges of AI data can provide substantial benefits, drawing on the deep well of existing governance frameworks.
One of the more contentious topics in my conversations at DGIQ West was defining the scope of AI. While large language models have captured much attention, AI encompasses machine learning, deep learning, natural language processing, and computer vision. The scope you assign to AI will inevitably shape your governance approach. A clear, consensus-driven definition of AI is essential for effective governance.
The scale at which AI operates requires a paradigm shift in our thinking, particularly in areas like testing. Testing AI systems is a complex and thorny challenge that we are not fully equipped to handle yet. This complexity underscores the need for rigorous testing governance and management frameworks, as well as test automation tools that can adapt to the unique demands of AI.
One practical and necessary first step in AI governance is the implementation of guardrails. Many organizations are hesitant to take this step, but it is crucial. Establishing clear policies, ensuring quality communication, and deploying a few simple technological guardrails can significantly mitigate risks. These measures ensure that AI is used appropriately, aligning with organizational goals and ethical standards.
Key Takeaway: Data contracts are a powerful and emerging tool in data governance, facilitating clear and accountable interactions between business and technology
An emerging but powerful tool in data governance is the concept of data contracts. At DGIQ West, the topic was addressed by just one presentation, delivered by SODA. Data contracts are a flexible and easy-to-implement pattern that can facilitate conversations between business and technology. They offer a structured way to manage data interactions, ensuring clarity and accountability.
Key Takeaway: Data Governance is willing to contribute to AI by being proactive and understanding key human and political factors.
Whether it was seasoned professionals or wide-eyed newcomers, data governance is hungry for ways to contribute effectively. This is challenging for many as they struggle with being seen as an inhibitor. Getting a seat at the decision-making table with early proactive activities is the target and people want to understand the human and political factors to influence well. My presentation about emotional intelligence was very well received and provided tools for attendees to engage their stakeholders with resonant leadership.
As AI continues to evolve, integrating these disciplines will be vital. By treating AI as a data product, leveraging the principles of data governance, and implementing practical guardrails, organizations can take clear, tactical steps to harness the full potential of AI while mitigating risks. If you want to leverage this in your organization and could use our expertise, contact us today.
