A new MIT Technology Review Insights report shows 82% of respondents are prioritizing data integration solutions that address more than just one specific use case. They want solutions that work across the board, last for many years and ensure compliance.

However, data governance and security concerns remain the second most common data readiness challenge. Stewart Bond, Vice President of Data Intelligence and Integration Software Services at IDC, cautions, “Organizations may have no control over someone using a piece of data in a business application and sending it to a generative AI model. These are critical concerns.”

Businesses in every industry are considering the impact of their data governance decisions, and the report, which surveyed 300 global C-suite and senior technology leaders, stresses the long-term need for better data governance management. 

[CTA_MODULE]

Data governance impacts AI implementation 

Mike Hite, Chief Technology Officer at Saks, explains that “when talking to boards or other leaders in a business, it’s often ‘outcome, outcome, outcome.’ But among technology executives, the first thing we always talk about as a group is governance.” Effective governance is the key to successful data management across any organization. 

Data governance isn’t just impacting current business outcomes though; it has the potential to affect how AI is implemented across organizations worldwide too. The MIT survey reveals that 60% of respondents agree their organization will need to rectify data governance, trust and security issues before it can achieve its AI goals. 

CDOs are on the frontlines of this evolution, with the study finding that 74% cite governance and security as primary challenges in preparing data for AI. In contrast, only 36% of CEOs and other C-level executives consider it a priority, highlighting a disconnect in perspectives and raising questions about how to manage and secure data while delivering necessary insights.

Governance concerns span across industries

While CDOs work to address trust and security issues, different industries harbor various levels of concern when it comes to data governance being a primary challenge of preparing data for AI. 

Concerns are especially prominent among government and financial services respondents, with 95% of government respondents and 70% of those in financial services calling governance and security primary challenges. 

‍

Concerns are less significant among respondents who are not in highly regulated sectors. But those individuals must be prepared to adapt as AI shakes up the data governance landscape. Otherwise, they risk being left behind by companies better prepared with stronger data governance frameworks. 

The intersection between data governance and GenAI

Many organizations are increasing the usage and spread of data throughout the company. But with many organizations using pre-built GenAI models, they must carefully consider how data is used and shared within and outside the organization. 

Regardless of size, businesses must be aware of the potential outcomes impacting both themselves and their customers when data is used in this manner. Businesses need steps in place to secure their data — they need solid data governance strategies. “The more you can have policies and control around data in an operational store, the better off you’ll be,” Bond says. 

Hite adds, “For Saks, we have a customer data governance framework in place that helps us ensure we are responsible with our customers’ data and use it to better serve our customers.” 

Building a robust data governance framework 

Building a solid data governance framework takes forethought and a secure data infrastructure, and the survey respondents agree, with 100% of government and 63% of financial services respondents including built-in security and governance among the most important attributes for data solutions. 

MIT’s research outlines several key factors for developing a data governance framework that is compatible with AI as well as both current and future technologies:

1. Put data governance first. No matter the industry, you need provisions in place now to manage and secure your data as it moves across the company. Doing so will increase data accessibility and visibility, improve accuracy and maintain quality — all while providing more targeted insights.
2. Anticipate security needs. As GenAI’s business impact grows, data workflows need to address security at the forefront — not as an afterthought — to protect the integrity and confidentiality of data as it flows around the organization.
3. Invest in powerful pipelines. Ensure your data has clear pathways to navigate and places to land — which becomes key if you’re looking to assemble a RAG architecture.  

Fivetran can help you as you lay a strong data foundation for AI. As the industry-leading data movement platform, we enable AI innovation for companies, such as Saks, Cemex and HubSpot, by providing automated access to centralized, cleansed and governed data.

[CTA_MODULE]