Rein in the work around the decision 

Much of the work required to close an NMTC transaction is the work surrounding the investment decision:  

• • Collecting documentation 

• • Reviewing third-party reports 

• • Reconciling information across multiple sources 

• • Managing compliance requirements 

• • Coordinating among project participants  

These activities demand significant staff time, even though many follow established processes that depend on consistency, careful review, and complete documentation. They may even be perceived as an obstacle to participation for those newly interested in NMTC opportunities. 

AI can improve these workflows by automating the organization and identification of the information that supports professional analysis.  

• • Documentation can be matched against transaction requirements as it is received, compliance checklists updated automatically, and project files organized throughout the diligence process.  

• • AI can also compare information across applications, legal documents, and third-party reports – to ultimately identify discrepancies or missing information that warrant additional review before a transaction advances. There is also an opportunity to leverage AI to pull data from more nontraditional datasets that are more time-consuming for community development personnel to mine on their own.  

This early analysis creates a more informed starting point for project teams, and a more consistent, efficient diligence process that directs attention to where it is truly needed. Professionals can begin their review with a clearer understanding of the transaction, the outstanding questions, and the areas that require judgment, without first spending hours locating documents, reconciling conflicting information, or assembling issues for follow-up. 

Better information access strengthens investment analysis 

Every NMTC investment begins with questions about the broader context: 

• • How have community conditions changed? 

• • Where have previous investments contributed to measurable progress? 

• • Where might future capital have the greatest impact? 

Answering those questions requires organizations to assemble and interpret information from multiple sources, often across many years.  

AI creates new opportunities to work with information organizations already possess.  

• • Public datasets, decades of census information, and historical portfolio performance can be analyzed together to provide a broader understanding of community conditions and investment outcomes.  

• • Organizations can explore relationships across datasets, compare similar census tracts, and identify patterns that merit further investigation.  

• • Historical trends may highlight communities where conditions have improved, areas where needs remain persistent, or geographies that share characteristics with previously successful investments. 

• • Combined with an organization's own portfolio data, these insights can support pipeline development, portfolio construction, and long-term investment strategy. They may also be helpful to supporting advocacy efforts, providing concrete examples of program potential or impact. 

Note, however: The data provides valuable context, but it does not answer every investment question. AI can help reveal that conditions changed within a community over time, but further analysis and professional judgement are required to determine why those changes occurred, including whether a particular NMTC-supported healthcare facility, rural development, educational investment, or other project contributed to that outcome. Investment teams remain responsible for interpreting AI-assisted information through local knowledge, organizational priorities, and an understanding of the communities being served. 

Governance: Protect progress with responsible adoption 

Successful AI adoption depends on governance that allows NMTC participants to improve operational efficiency while preserving the rigor, accountability, and community focus central to the program. 

Responsible AI governance should give organizations a practical framework for using the technology effectively and safely. An effective AI policy: 

• • Establishes approved tools 

• • Defines appropriate applications 

• • Clarifies expectations for reviewing AI-generated work 

• • Sets standards for data quality and privacy, including what information should never be entered into AI platforms (e.g., consider federal, state, and local restrictions on Personally Identifiable Information (PII) like Social Security numbers) 

• • Gives employees practical guidance for using AI productively within these established safeguards 

A closer look at some of these considerations: 

Policies should be straightforward enough for employees to apply consistently, yet flexible enough to evolve with changing technology and organizational experience, as well as any relevant regulation or industry guidance that may have changed. AI governance is not a one-time exercise; it develops as capabilities, risks, and organizational practices continue to mature. 

Governance also extends to technology selection. Directing employees toward vetted enterprise tools – and limiting unapproved applications – helps protect financial information, project documentation, and other sensitive data. Before integrating AI platforms into operational workflows, organizations should evaluate them carefully for security, privacy protections, regulatory compliance, and the way information is stored and managed.  

Historical data requires its own form of scrutiny, against bias. AI models learn from existing information, but historical investment patterns do not always reflect historical opportunity. A rural community that received limited investment in the past, for example, may appear less attractive when viewed solely through historical lending data, even when the underlying need remains significant. Organizations should evaluate AI-generated analysis with an awareness of these limitations and continue applying professional judgment when assessing community needs and future investment opportunities. 

Effective governance extends beyond data security and regulatory compliance. It gives organizations a structure for adopting AI deliberately, evaluating outputs critically, and refining policies as the technology and its applications evolve. 

How to adopt AI: Support ongoing development toward organizational fluency 

Responsible AI adoption depends on organizational learning. New capabilities are constantly emerging, so AI should be a recurring discussion. Regular conversations about new tools, practical applications, implementation challenges, and lessons learned help leadership stay informed while encouraging ideas to emerge from across the organization.  

For instance, when deploying AI, organizations should provide employees with a controlled “sandbox” environment to test approved tools, compare models, refine prompts, and understand how AI performs – before new capabilities enter production workflows. This unstructured experience often reveals practical opportunities, limitations, and governance needs that policy development alone cannot anticipate.  

In complement: Pair unstructured with structured activity. Comparing the same prompt across multiple models, or refining a prompt within the same model, develops a practical understanding of how AI behaves. These exercises demonstrate how context, wording, and previous interactions influence the quality and direction of a response – which helps users evaluate outputs, ask more precise questions, and recognize where additional review is required. Always remember to double-check the answers provided and ask where the data was cited from. 

Finally, draw small groups from across the organization to identify practical opportunities for AI across the transaction lifecycle: Investments, compliance, operations, asset management, technology. Exploring those use cases together helps organizations build a unified experience while refining governance, policies, and operational practices. 

In conclusion: Enhance your processes with strategic AI investment 

AI capabilities will continue to evolve, requiring organizations to keep learning, testing, and refining their approach. Clear governance, practical experimentation, and shared organizational knowledge can strengthen investment and compliance processes while helping NMTC participants evaluate new applications responsibly. 

Reach out to discuss how AI can best support your NMTC strategy – and how to make sure your organization supports AI success.