I recently explained the significance of data management as an enabler of strategic adoption of artificial intelligence. Data management enables enterprises to ensure that data is valid, consistent and trusted for operational use cases and analytic decision-making. Large volumes of data are required to train models, making data management and data governance critical to AI. Data quality and data integration issues have the potential to prevent enterprises from delivering on expectations. As the recent 2025 ISG Buyers Guide for Data Management highlighted, AI is also critical for enabling enterprises to manage the volume and variety of data.
Data management includes data governance, data quality, master data management, data integration and data intelligence. Products addressing these capabilities ensure that
enterprises collect, store and process data in accordance with strategic goals and regulatory requirements. The expectations placed on AI initiatives by senior executives are raising awareness of the importance of data management. Data quality and data integration issues have the potential to prevent enterprises from delivering on expectations. More than one-half (54%) of participants in ISG’s Data Market Lens study cited data usability for AI as a data challenge, while 40% cited data siloes or fragmentation. While AI-ready data is clean, well-organized and compliant with regulatory standards, too many enterprises find themselves struggling with data that is fragmented, inconsistent and not easily accessible.
Traditional approaches to data management relied on manual functionality and the expertise of data stewards. Traditional techniques and expertise, however, are no match for the complexity of data sources and performance requirements involved with delivering real-time AI applications. More than two-thirds (70%) of participants in ISG’s Data Market Lens study cited the inability to expose or share data due to inefficient or ineffective tools and platforms as a primary disruption to data initiatives. As a result, many enterprises are seeking data management products and tools infused with AI to automate key data management functions.
Data management tasks are prime candidates for automation and acceleration using AI. They are often repeatable, routine and mundane, but also time-consuming. Examples include searching for data, profiling and tagging data, data matching, cleansing and standardization, along with creating and implementing data quality rules and documenting data integration pipelines. Automating these tasks with AI can enable data management experts to improve productivity and focus on higher-value tasks that more directly support business goals and innovation. Justifying the investment in tuning AI models and creating applications to automate data management tasks might be difficult, however, since the tasks themselves are typically non-differentiating. As such, many enterprises are looking to software providers to deliver packaged functionality to automate data management.
The 2025 Data Management Buyers Guides illustrate that AI for data management is a work in progress for many software providers, with significant variation between providers in terms of functional maturity. The greatest progress has been made in adopting generative AI to enhance data discovery and data descriptions. More than two-thirds (70%) of providers assessed in our Buyers Guide for Data Intelligence graded A- or above for the use of natural language search interfaces for data discovery, while 60% graded at the same level for the use of AI to automate data asset descriptions. At the other end of the spectrum, only 10% of providers assessed in our Buyers Guide for Data Governance graded A- or above for the use of AI to automate the masking or obfuscation of sensitive data, while only 16% of providers assessed in our Buyers Guide for Data Quality graded at the same level for the use of natural language to define data quality rules.
Data quality software providers are more advanced in the use of AI to automate and enhance data quality checks, for which more than one-half (56%) of providers graded A- or above, while 52% graded at the same level for the use of AI to automate and enhance data profiling. Another key investment area for AI is data integration. Almost one-half (47%) of providers assessed in our Buyers Guide for Data Integration graded A- or above for natural language data integration pipeline creation, for example, while 44% graded at the same level for providing recommendations to enhance data integration pipeline design. Automating the documentation of existing data integration pipelines is another potential use case, but it is relatively immature, with only 38% of providers assessed grading A- or above.
The least mature product category in terms of investment in AI is master data management. Although more than one-half (58%) of providers assessed in our Buyers Guide for Master Data Management graded A- or above for the use of AI to automate and enhance data matching, other areas are clearly at the early stages of functional development. Only one-quarter (25%) of providers assessed graded A- or above for the use of AI to automate and enhance data merging, while just 17% graded at the same level for the use of AI to
automate and enhance data cleansing. Other areas that show some room for improvement include the use of AI to automate and enhance data modeling (17%), recommendations for data usage (19%), data enrichment (33%) and data integration pipeline error resolution (34%).
I assert that through 2027, almost all data intelligence software providers will deliver support for GenAI-driven assistants to automate and accelerate data management and integration processes. I recommend that enterprises evaluating data management software pay close attention to the details of AI functionality to ensure that it matches expectations in terms of the potential to automate data management tasks and accelerate success with AI. Many software providers now offer copilot functionality to deliver generic assistance by accessing product documentation and best practices information to generate answers to user questions. While this is useful, it is not as beneficial as functionality that has been designed and tuned to automate the fulfillment of a specific task.
Regards,
Matt Aslett
Fill out the form to continue reading.