Where information governance meets machine learning.
It’s an understatement that the ever-increasing volume and velocity of every imaginable type of data — text, image, map, video, etc. — has been a major force driving artificial intelligence and machine learning across next-generation applications. From robotic process automation that deploys smart bots in the enterprise to learn and automate repetitive tasks, to social/consumer applications that continuously capture and learn from the behavior of users, AI is exploiting the true power of data to deliver a new level of automation and efficiency.
Although AI was conceived decades ago, for the first time it is tangible within applications we see and use every day. And with the pace at which data is being created and stored, the use of AI to learn and predict events, actions, and decisions is only going to accelerate. Indeed, these are exciting times to leverage machine learning and AI.
EXAMINING DATA GOVERNANCE
However, if data is at the very foundation of this AI revolution, we need to analyze samples to understand how data is used as part of a new genre of AI-driven applications. This involves answering some important questions. What is the role of data governance and data quality in a world of real-time automaton driven by machine learning and AI? How does this affect the bedrock principles of governance, such as who has authorized access to data, where the data is stored, how the data is used and applied, and how long it is retained and archived? When manufacturers have critical manufacturing lines being driven in real-time by AI, how will an organization ensure that its use of data adheres to industry and legal regulations, or even a code of ethics?
As the volume of data increases and the breadth and usage of AI expands, the role of information governance could become even more critical to organizations that see and use data as a core asset. Without question, it will expose the need for next-generation governance platforms to move beyond a documentation exercise that merely catalogs terms and policies for centralized user access. A document- and glossary-centric approach will not be agile enough to address the increasing complexity of how data is driving the digital enterprise.
INTEGRATING AI WITH GOVERNANCE
Data governance boards and chief data officers should be keen to look for innovative governance platforms that include AI functions as part of their core offering. This will allow them to make smarter decisions on how and where all types of data are used across the organization. Moreover, they should look to platforms that implement, integrate, and execute governance where AI lives — at the system and application infrastructure level.
When governance is implemented in this way, it will evolve from a manual administrative exercise to a data-driven approach — one in which policies defined and documented by data analysts are not static and isolated, but rather dynamic and combined at the data integration layer. With this approach, data integration will incorporate meta data, system data, application data, business rules, and process context that can leverage the advancements in machine learning and AI in order to continuously learn and integrate policy as part of an enforced governance infrastructure.