AI Set to Disrupt Traditional Data Management Practices

While strong data management has long been a foundational practice for business intelligence and analytics, enterprise organizations will need to update what they do to meet the needs of growing advanced analytics implementations such as machine learning and other artificial intelligence in the years ahead.

That was one of the conclusions of Gartner’s recent Data and Analytics Trends report for 2022.

“Without the right data, building AI can be risky and even dangerous,” says Gartner VP Rita Sallam. “Most organizations, from a data management perspective, address very important AI-specific considerations for data management like data bias, diversity, labeling. They often address those things haphazardly.”

That’s because the AI community is focused largely on model development.

Why Data-centric AI Will Disrupt Data Management

“They often don’t have the skill sets, or their organizations don’t put in place processes and tools and practices to really manage data management for AI specifically,” says Sallam. “So data-centric AI has the potential to disrupt what has been traditional data management practices as well as prevalent model-centric data science by making sure that AI-specific considerations like data bias, labeling, drift, are all in place in a consistent manner to improve the quality of models on an ongoing basis.”

Are tools under development to address this need, or are organizations investing in solutions for it? Sallam says that some of the other trends on the list will contribute to improving data management around AI.

Specifically, to address this gap, leading organizations are disrupting data management for AI by building out data fabrics on active metadata and investing in things like AI governance, she said.

This data-centric AI trend is one of several Gartner highlighted in its report for 2022 and grouped with a few others under the title of activating dynamism and diversity. It leads into another one -- metadata-driven data fabric -- which feeds into that disruption of data management for AI.

“Metadata is data in context. It’s the when, the where, the who, the how aspects of data,” says Sallam. “Data fabric listens and learns and acts on that metadata and then applies continuous analytics over existing and discoverable and inference metadata assets.”

Other Data and Analytics Trends

Other trends in this group are adaptive AI systems that allow organizations to adjust quickly to innovation and the movement to always share data rather than enabling an internal culture of data hoarding.

Gartner says that the trend to “always share data” reinforces data sharing as a business-facing key performance indicator. It helps achieve effective stakeholder engagement and increase access to the right data to generate public value. As an example, Gartner cited the COVID-19 pandemic as one where the sharing of data helped accelerate independent and interrelated public and commercial digital business value -- from setting public health guidelines to developing vaccines.

These four trends are all interrelated, Sallam says, because you need data fabric as the foundation with metadata providing automatic discovery of data assets and services.

Another one of the trend groups is “Augment People and Decisions,” an area that includes prioritizing data literacy across the workforce, which is key for organizations with strategic data-driven business goals. This group of trends also includes context-enriched analysis, business-composed data and analytics, and decision-centric data and analytics.

What is Connected Data Governance

The final group of trends is collected under the umbrella of “Institutionalize Trust,” and includes connected governance, AI risk management, vendor and regional ecosystems, and expansion to the edge. This group relates back to all the other trends, too.

Gartner analyst Ted Friedman says that the idea of data governance has gotten harder because organizations have gotten more complex and data and analytics activity is happening in a more distributed fashion.

“We see siloed or point governance-related initiatives running, but they are dissimilar. They’re not standardized. They are not leveraged or aligned in any way,” he says. “This is what our thought of connected governance is about. We’re seeing a very important trend evolving where organizations are trying to link together the variety of governance-related initiatives that are happening across their business.”

These initiatives include getting key stakeholders involved including the chief information security officer (CISO), as well as data and analytics leaders who are focused on data governance from a quality and lifecycle and master data point of view. Friedman says that organizations shouldn’t just be talking about data governance. They should be talking about data and analytics governance. That’s because governance needs to incorporate artifacts related to AI, things that operate on data, or derivatives of the data, he says.

All these trends are interrelated as data and analytics practices evolve in organizations and grow more complex. Organizations should keep an eye on all of them, but focus on what are priorities for their own enterprises.

“The top data and analytics trends and technologies don’t exist in isolation,” says Sallam. “You’ll see that they build and reinforce on one another.”

What to Read Next:

Enterprise Guide to Data Privacy

What You Need to Know About AI Ethics

AI’s Bumpy Road to Fruition