The Department of Defense, through the Air Force District of Washington (AFDW), is seeking proposals for the Small Business Innovation Research (SBIR) project titled "Securing Legacies: Developing an AI-Enhanced Privacy Architecture for Military Family Knowledge Systems." The objective of this initiative is to create a privacy-preserving architecture that securely collects and correlates personal data from various sources, ensuring access is granted strictly on a need-to-know basis while addressing critical privacy concerns related to personally identifiable information (PII) and sensitive health data. This project is vital for supporting military families by establishing a dynamic knowledge base that integrates data from the Department of Defense, Veterans Affairs, and other agencies, ultimately enhancing the sense of belonging and purpose for service members and veterans. Interested parties should note that the application due date is November 6, 2024, and further details can be found at the solicitation agency's website: https://www.dodsbirsttr.mil/topics-app/.

## Description

This research aims to develop a privacy-preserving architecture that includes both information and technical structures to ensure the secure collection and correlation of personal data from multiple sources, with access based strictly on a proven need-to-know basis. This architecture encompasses several critical aspects:<ul><li> Anonymization: Data subjects will be anonymized, with identities only revealed when absolutely necessary, such as for making contact for a specific purpose.</li><li> Compartmentalization: Only the minimal information necessary for a given purpose will be disclosed, whether the subject's identity is revealed or remains anonymous. For instance, instead of disclosing a subject's exact Date of Birth, the system may simply verify that the person is over the age of 21.</li><li> Lineage and Provenance: There will be clear traceability between source data and derived information to ensure that the origins and transformations of data are comprehensively documented and transparent.</li><li> Consent: Appropriate consent must be obtained, documented, and linked to the data at both the entity and individual attribute levels as required.</li><li> Access Control: Access controls will be meticulously implemented not only at the individual level but also down to specific data attributes. This research will investigate methods to enable such compartmentalized access while maintaining the integrity and completeness of the entire system.</li></ul>AI should be leveraged to strengthen the privacy-preserving architecture by identifying policy mismatches and preventing the exposure of personal information. For example,<ul><li> Policy Compliance Monitoring: AI algorithms will continuously review data handling and access protocols, automatically flagging deviations from privacy policies for swift correction.</li><li> Sensitive Information Exposure Detection: AI will monitor patterns and access logs to preemptively detect and mitigate unauthorized access to sensitive data, reducing the risk of breaches.</li><li> Real-time Risk Assessment: AI will constantly assess risks to data privacy and security, identifying potential vulnerabilities and suggesting proactive mitigation strategies.</li></ul>

## Objective

Supporting Air Force and Space Force families and preserving their legacies is a crucial mission of the Air Force District of Washington (AFDW). To achieve this, we need a dynamic knowledge base that integrates data from the Department of Defense (DOD), Veterans Affairs (VA), and other agencies, and potentially augmented by publicly available information on social media platforms. This system will continuously update service member, veteran and military families areas of interest or concern from the inception of their service and employ predictive analytics to anticipate family goals and needs and manage skills of these service members and veterans as a vital national asset, and possible ways to communicate during times of national emergencies. Furthermore, it will improve how military units and memorials are able to connect to military and non-military audiences to educate them on the military&rsquo;s history and both organizational and individual contributions to national defense. However, the extensive data collection and correlation required to establish such a knowledge base necessitates that privacy and confidentiality concerns are addressed with utmost seriousness. The data involved includes personally identifiable information (PII) and sensitive health information potential protected by Health Insurance Portability and Accountability Act (HIPAA) &mdash;both crucial for effectively serving our members. Moreover, the potential for elevated classification levels or increased sensitivity categorization due to data aggregation, especially the correlation can now be automated using AI, must be rigorously managed. At the same time, AI can be leveraged to strengthen the privacy-preserving architecture by identifying policy mismatches and preventing the exposure of sensitive information, which will help safeguarding this information in military family legacy systems and other governmental operations.The outcomes of this research will establish robust protocols for managing sensitive information, thus enhancing the security and privacy of data within military family legacy systems and potentially other governmental data systems. While at the same time enabling an AI enabled sense of belonging and purpose for those who have or continue to serve, despite changes in career, location, or time. 

## Phase I

Phase I work will establish the feasibility of a privacy preservation architecture. Specifically, it will develop a reference architecture for a legacy knowledge base, detailing key components, their interactions, and the open technology standards applicable for implementation. Additionally, Phase I will demonstrate how information can be protected using the reference architecture, using a use case co-developed with AFDW. And provide a next generation social network for those who serve.

## Phase II

Phase II will develop a prototype that integrates data from various Department of Defense (DOD) and government systems, augmented by public and social media platforms to demonstrate how this integrated data can offer a comprehensive profile of an individual. This phase will showcase the use of AI to proactively anticipate the person&rsquo;s needs. Additionally, it will illustrate how privacy is maintained by displaying different levels of information about the same individual, tailored according to the user's specific purpose and need-to-know basis.

## Phase III: Dual Use Applications

The technology has versatile applications across both Department of Defense (DOD) systems and commercial sectors, wherever personal information is collected and analyzed. Beyond personal data, its utility extends to any system where higher classification information might be derived from aggregating multiple sources of lower classification data. This broad applicability underscores the technology's potential to enhance security and privacy across diverse platforms.

1730145487909.pdf

The document outlines a project, titled "Securing Legacies," aimed at developing an AI-Enhanced Privacy Architecture for Military Family Knowledge Systems under the Air Force District of Washington (AFDW). The objective is to create a dynamic knowledge base integrating data from the Department of Defense and other agencies to support military families. The architecture will ensure the secure collection of personal data while addressing privacy concerns, particularly concerning personally identifiable information (PII) and sensitive health data.

The proposed research includes essential features such as data anonymization, compartmentalization, lineage tracking, consent documentation, and meticulous access controls. AI technology will play a vital role in monitoring privacy compliance, detecting unauthorized access, and assessing risks to enhance data protection.

The project is structured in three phases: Phase I focuses on the feasibility and reference architecture for a knowledge base; Phase II involves developing a prototype that integrates data from various systems; and Phase III explores dual-use applications of the technology across both DOD and commercial sectors. Ultimately, the initiative aims to establish robust protocols for safeguarding military family legacies while enhancing the sense of belonging for service members and veterans.

Topic

Agency

Program

Topic Link

Additional Information