DoDIIS Worldwide Conference 2023 Hotels

Breaking Barriers with Generative AI: Enhancing Systems Security and Data Sharing for the Warfighter

In the presentation “Breaking Barriers with Generative AI: Enhancing Systems Security and Data Sharing for the Warfighter,” we will explore the transformative potential of Generative AI in the context of emerging technologies to support the warfighter. This presentation directly addresses the theme of the conference, which focuses on the intersection of systems security, access management, and data sharing.

The Department of Defense (DoD) should care about the application of Generative AI because it offers a unique opportunity to overcome existing barriers and enhance the DoD’s systems security and data sharing capabilities. Generative AI has the power to revolutionize the way the DoD operates by enabling the creation of synthetic data, generating realistic scenarios, and simulating complex environments. This technology can significantly improve training, testing, and decision-making processes, leading to more effective and efficient warfighter operations.

By leveraging Generative AI, the DoD can enhance systems security by simulating and identifying potential vulnerabilities, predicting and countering cyber threats, and developing robust defense mechanisms. Additionally, Generative AI enables secure and controlled data sharing, allowing the DoD to collaborate with partners, share information across agencies, and leverage collective intelligence while maintaining data privacy and integrity.

The impact of embracing Generative AI in the DoD environment is significant. It empowers the warfighter with advanced tools and capabilities, enabling them to make informed decisions, respond rapidly to evolving threats, and achieve mission success. By breaking barriers with Generative AI, the DoD can enhance its operational effectiveness, improve situational awareness, and ultimately ensure the safety and security of the nation.

What can we do about AI?

Leveraging AI to augment our information forces gives us massive new capabilities. Adversaries know the same thing and are trying to do the same thing. A small amount of high-performance computing (HPC) in the right places will solve many problems of AI relating to deployment, engagement, and data ingestion in environments where data security and access controls are paramount.

Using AI in secure, reliable, resilient, rapidly updated ways will give us an edge. Relying on commercial cloud providers for all computing, R&D, and services for machine intelligence is a risky way to get that edge. Relying on commercial cloud for the foundations and using in-house HPC expertise and resources to deliver the last mile of machine intelligence will reduce risk and accelerate the adoption of secure, reliable, robust, and repeatable AI inside the enterprise.

Automation and Specialized AI to Accelerate Warfighter Impact

Today’s warfighter is more connected than ever before to a streaming vector of actionable intelligence. Platforms, systems, and data – all traversing an ever-increasing number of endpoints. As we look to events around the world as leading examples of how the battleground continues to change, we are called to action to improve both the offensive and defensive digital capabilities of our military. To win, our priorities must clearly align to automating heterogeneous environments at a moment’s notice, delivering consolidated AI-infused digital experiences to each warfighter, and leverage Automation and AI to protect our digital advantage.

Qrypt Technology Mitigating the Quantum Threat to Secure Communications

Scaling quantum computers will eventually break the digital security used in virtually all modern data networks. For decades, our adversaries have been collecting encrypted communications with the intention of decrypting and operationalizing it when larger quantum computers become available. This Cold War technique is known as “harvest now, decrypt later” (HNDL); it makes headlines today because quantum computers can break our existing algorithms by brute force. The transition to Post Quantum Cryptography (PQC) does not solve the HNDL problem because the new algorithms have no mathematical proof of hardness. As such, NIST advised developers to be “crypto-agile” and prepared to replace PQC at any time in the future.

For decades, implementation errors, weak encryption keys, poor randomness, corrupted software libraries and a variety of attacks resulted in the total exploitation of stored HNDL data. The issue is fundamental to the single-points-of-failure in public key infrastructure (PKI) which is based on a 1970s architecture predating the internet, cloud, virtualization, and containerization used in modern information systems. Qrypt leverages multiple quantum entropy hardware sources and distributed software algorithms to enable end-to-end-encryption (EE2E) with simultaneous key generation at any endpoint. This mechanism decouples the data from the decryption keys, eliminates key distribution and is unaffected by multiple weaknesses in the system, including the potential failure of the PQC algorithms and insider threats.

The modern warfighter will operate in converged PKI environments on 5G/6G networks, using autonomous systems, in smart cities, built on technology under adversarial control. Secure communications will need much higher levels of assurance than currently possible. Incremental improvements to classical techniques will be insufficient in the quantum era.

Containerized Digital Engineering Applications Through Multilevel Security Architectures

Scott Devitt, GDMS
Brian Newson, GDMS

The GDMS Chief Engineer for Multilevel Security, Scott Devitt, will demonstrate and explore real-world SAP use cases with MLS containers for DE Environments. During his 37 years with General Dynamics, Scott has designed, built, installed, and maintained classified capabilities for the DoD and IC including operational mission cells supporting forward locations with multiple stove piped networks at different classification levels. His presentation will highlight the value of a DE polyinstantiated or containerized framework in safeguarding SAP data and the benefits of leveraging a multilevel file share when working across multiple connected classified environments. It will also discuss the challenges faced in integrating the innovative capability into legacy stovepipe SAP networks with existing applications and explore potential solutions.

In summary, these three leading edge MLS DE design patterns present a robust set of solutions to the growing challenge of collaborating and working effectively in the ever-complex SAP community. By leveraging this capability, organizations can bolster security, consolidate costly licenses across networks and safeguard their most valuable data while also dramatically improving user operational efficiency on their primary network. By employing containerized applications, data transfers between networks are eliminated, reducing the risk of information leakage through unauthorized channels.

DoD Supercomputing for Special Programs and SCI Projects

Kelly Dalton, AFRL
Jonathan Thompson, AFRL

This is an update to last year’s presentation regarding an effort to provide DoD funded, shared supercomputing to the acquisition engineering, research, development, and test & evaluation communities. Large scale supercomputers are funded by the DoD High Performance Computing Modernization Program for the purpose of providing no-cost computing to scientists and engineers working on DoD problems. Contractors can also access these resources under a DoD contract involving an RDT&E project. This unclassified/CUI presentation will provide information regarding current status and future plans by the Department of Defense to provide continued access to free supercomputing resources to government and contractors supporting special programs and/or SCI-related projects in the research, development, acquisition, and test & evaluation mission areas. Specifically, the large-scale computing resources provided by the DoD High Performance Computing Modernization Program (HPCMP) will be discussed as well as how to access these resources. The supercomputing systems undergo a recurring technical refresh funded by the DoD HPCMP. The individuals/organizations do not pay for compute time or storage on the DoD supercomputers as these are funded through the DoD HPC Modernization Program.

Detection of Adversarial FOCI in Open Source Software

USG has prohibited acquisition of hardware from sanctioned entities and excluded those companies from doing business in the United States. But most program managers don’t realize that those same sanctioned entities and foreign adversaries actively develop, maintain and control software dependencies used by classified military programs. While these dependencies can theoretically be code-reviewed before approval, they’re almost never reviewed beyond a one-time check for viruses or known vulnerabilities – with little to no monitoring of upstream risks. And even if their source code is reviewed, there’s no chain of trust between repositories and published packages.

This talk will illustrate how Chinese and Russian developers are positioning in the upstream software supply chain, how that risk can be detected and how it can be managed in an automated way, at scale, in the absence of any known or detectable vulnerabilities in the code. Single-maintainer projects belonging to Russian government employees have been identified in federal APIs that handle highly sensitive data at high scale. The ecosystems in which adversarial entities are active include AI/ML used in defense, which was the subject of a year-long analytic project that Ion Channel (recently acquired by Exiger) executed for DTRA. The data backplane for identifying adversarial FOCI in upstream software dependencies has both defensive and offensive value in software-intensive programs and missions.

How AI Can Support Data Security in the Cloud

Kathleen Featheringham, Maximus
Michael Sieber, Maximus
Frank Reyes, Maximus

As the Defense Department (DoD) continues its cloud modernization journey with the Joint Warfighting Cloud Capability (JWCC) and other programs, managing sensitive data in the cloud is a top priority and cybersecurity challenge.

Emerging technologies such as artificial intelligence (AI) offer novel strategies to fortify cryptographic practices, enhance data encryption, and bolster cloud security.

Reaping the benefits of AI-powered cloud security requires good data practices and data governance as well as proper configuration management and modern encryption strategies to ensure data security.

This session aims to address common cloud security concerns and outline use cases for comprehensive cybersecurity and encryption practices powered by AI to properly manage sensitive data in the cloud.

Building The Secure Future: Large Language Models & Secure Classification Review

Jack Gumtow, MetroStar
Cedric George, MetroStar
Joe Early, MetroStar

With decades of sensitive government documents requiring classification review, manual human intervention has been vital but cumbersome. While some technologies aid the process, none—to-date—reduce the cognitive load of a human reviewer. DeCypher (an AI solution) employs cutting-edge language models and classification guidebooks, delivering swift, explainable, and automated classification, markup, and redaction with the essential oversight of human expertise. This talk aims to discuss how we can embrace the future of secure and efficient document classification by diving into the game-changing innovation of AI and Large Language Models.

Insights Unveiled - How Blockchains Are Shaping Our Warfighter’s Digital Future

Derek Claiborne, Chainalysis
Brian Smith, Chainalysis

Web3 is all about innovation and collaboration – but with that comes heightened risks. Chainalysis has a commitment to creating a safer environment for all who enter the world of Web3. In this discussion, we will explore blockchain’s potential in addressing challenges faced by our warfighters. The evolving threat landscape involving strategic competitors, rogue nations, and terrorist groups is examined, with a particular emphasis on their exploitation of cryptocurrencies for illicit activities. The role of blockchain technology in countering these threats is then elucidated, showcasing its characteristics like decentralization and transparency.

This includes a deep dive into using blockchain for geolocating threat actors and tracking illicit activities. International collaboration and the integration of blockchain-based intelligence into defense strategies are discussed as well. Challenges, considerations, future prospects, and recommendations for blockchain adoption in cybersecurity and defense form vital segments of the discourse, ultimately underlining the significance of embracing emerging technologies like blockchain to empower warfighters and enhance national security in an ever-evolving digital landscape.
Audiences will gain a comprehensive understanding of how blockchain technology can effectively address blockchain-enabled threats and enable the geolocation of threat actors in the realm of cybersecurity and defense. They will also recognize the pivotal role of international collaboration and blockchain integration in bolstering national security efforts across evolving global challenges.

Zero Trust Network Access 2.0

Mark Reyero, Palo Alto Networks

Delivering zero trust at an enterprise level begins with a fundamental change in how the DoD builds its cyber security architecture to prioritize both security and performance. Zero trust is built upon the foundation that there is already a malicious actor or compromised data or devices within the enterprise. This assumption creates a need for a process of continuous validation of users, devices, applications, and data in an entirely controlled and visible manner. With Palo Alto Networks’s Zero Trust Network Access 2.0 (ZTNA 2.0), coarse-grained access controls based on an “allow and ignore” model is left behind to introduce a consistent least-privilege access control model focusing on application layer security inspection.

Defense Technical Information Center Overview

Yvette Jacks, DTIC
Christopher Hunt, DTIC
Matt Poe, DTIC

Aligned to OUSD(R&E), the Defense Technical Information Center (DTIC) is the DoD’s central S&T knowledge and analytics center. DTIC’s Deputy Administrator, Chief Data Officer and Chief Technology Officer will provide an overview of DTIC, some exciting AI and ML capabilities we are building out, the ROI of DTIC’s centralized capability, and how the DoD S&T community can engage and partner with DTIC. Attendees will also learn about DTIC’s digital modernization journey, experience implementing DevSecOps, and gain insights into DTIC’s pathfinder experience working with DoD CIO to achieve the Department’s second ever continuous authority to operate (cATO).

Outprocessing the Checklist

Rob Case, DON SAP CISO

An examination of the Risk Management Framework as a dynamic cybersecurity program featuring Cyber Hygiene, Cyber Readiness, and Continuous Compliance as prime disciplines. The end in mind is to finally mature beyond checklists and firefighting, develop locally relevant threat intelligence programs, prepare for continuous ATOs with fully developed ConMon programs, and generate feedback loops between the monitors and responders. This presentation explores the Risk Management Framework and JSIG control families as features of Cyber Hygiene (management of the authorized) and Cyber Readiness (management of the unauthorized) and encourages RMF practitioners to go beyond the ATO. The concept of outprocessing the checklist is encapsulated in a change of mindset; completing a task is not compliance and compliance is not security. Narrative-based bodies of evidence authored and informed by ISSOs are insufficient. Cybersecurity practitioners must seek system-based artifacts as their proof of configuration and ISSOs must be informed by the system.

Enabling Human-Machine Collaboration with Visual Analytics

Chad Steed, ORNL

Visual analytics is a viable approach for enabling human-machine collaboration in today’s most challenging data analysis scenarios. While the increasing volume and complexity of modern data sets severely limits the viability of purely manual, human-centered strategies, most data analysis tasks are inherently exploratory (meaning the user doesn’t know all the questions they may ask of the data beforehand) and require interactive query capabilities. Visual analytics solutions that balance human and machine strengths are ideal, but achieving such a balance is not trivial. It requires judicious orchestration of human strengths, namely creativity, intuition, visual perception, and cognition, with the computational power of machines and the automated algorithms that run on them. In this talk, I will discuss modern data analysis challenges and how visual analytics tools can help solve them. To illustrate these ideas, several visual analytics systems will be described with an emphasis on the integration of human interaction, data visualization, and algorithmic guidance into flexible tools. I will also highlight the application of these tools to real-world applications in explainable AI, sensitivity analysis, multivariate analysis, and text mining. I will conclude with an overview of active and future visual analytics work.

AI Defined Networking

Caleb Snow, WWT
Kimberly Haines, WWT

AIDN leverages state-of-the-art machine learning and artificial intelligence algorithms to detect and respond to even the most advanced and elusive threats. It identifies malicious activities in real-time, minimizing the potential impact of attacks. Through continuous monitoring through AIDN, your organization enjoys 24/7 monitoring of your digital infrastructure. AIDN provides immediate alerts and proactive threat remediation to prevent breaches before they occur.

Through our User-Friendly Interface, our intuitive, user-friendly dashboard simplifies the complexities of cybersecurity management. It offers real-time insights into your network’s security posture, allowing for informed decision-making. AIDN is designed to grow with your organization. AIDNs threat intelligence integrates threat intelligence feeds from multiple sources, ensuring you stay ahead of emerging threats. This knowledge helps AIDN adapt its defenses and protect your organization from new attack vectors.

Intelligent Physical Infrastructure Solution for Zero Trust

Mohammad Raza, Green Lambda

Network infrastructure has to allow the administrators to understand, see, and monitor exactly where threats can enter the network. Add to this, intrusion detection, the ability for intrusion prevention and you have a network ready for government level security.

Mission Collaboration - Bringing the Power of Government and Industry Together with their Data

Kenny Bowen, Microsoft
Rebeka Melber, Microsoft

Historically, the DoD SAP Community has faced a glaring challenge – one of disconnection. Over the past decade, a remarkable transformation has taken place. Thanks to a roll out of enterprise-level SAP capabilities over the past decade, connectivity has surged to unprecedented levels. These advancements have become the backbone of an entirely new era, opening doors to a consolidated stream of data that is poised to reshape the landscape of national defense. The proliferation of Cloud Service Providers (CSPs) authorized for SAP data further signals the dawning of this transformative era. In the midst of this technological evolution, it’s crucial not to overlook the basics. While the buzzwords of Artificial Intelligence and Machine Learning are reshaping our technological landscape, the foundation for these innovations must be steadfastly established. Our success hinges on getting the fundamentals right, ensuring that the most fundamental functions are in place. Collaboration emerges as the cornerstone that will pave the way towards a truly robust and effective national defense strategy. This talk will delve into the narratives of the past, the dynamic landscape of the present, and the exciting potential of the future. It encompasses communication between Defense Industrial Base (DIB) and Government, Enterprise and Mission Users, and General and Privileged Users. As we stand on the precipice of unparalleled technological advancements, it is our responsibility to steer this transformation with clarity, unity, and a shared vision. Through collaboration and convergence, we shall not only bridge past disconnects but also construct a foundation for a stronger, safer, and more technologically empowered future.

All Components Left Behind: Assessing the Troubling State of our Technology Supply Chain

John Loucaides, Eclypsium

Not a month goes by without another deep vulnerability in CPUs, memory, BIOS, BMCs, or some other component buried inside nearly every piece of IT equipment. While these issues sound serious, the very premise of these components is to abstract away hardware details. With adversaries known to be exploiting these bugs, how can we assess vulnerabilities not mitigated by traditional endpoint security solutions?

In this talk, John will explore some of the technical issues related to cyber security of the supply chain. He will explain the most common issues, how to check for them, and how to avoid being taken by surprise. Having personally been involved in research into and coordinated disclosure of serious platform-level vulnerabilities, John will speak from personal experience (both within USG and outside) to suggest practical solutions involving both open source and commercial tools to help with this evolving problem. After discussing issues that affect firmware updates, end of life, component vulnerability scanning, integrity checks, and sanitization/destruction, attendees will discover that even though perfection is impossible, all is not lost.

Spill Coffee Not Data: How to Secure Sensitive Assets and Avoid Being Roasted by Adversaries

Andrew “AJ” Forysiak, Varonis
Chad Mason, Varonis

The United States faces persistent and increasingly sophisticated malicious cyber campaigns that threaten the public, private sector, and ultimately, the American people’s security and privacy. By implementing Zero Trust (ZT) across all agency systems, the U.S. government seeks to protect high-value assets, but without first building a solid foundation, any zero-trust architecture will be largely ineffective and unwieldy. Agencies must now strive to provide best-in-class zero trust-based security while satisfying compliance requirements such as EO 14028, DOE O 471.1, and OMB 22-09. Zero Trust represents a paradigm shift in how we think about protecting our assets and requires a multi-phased process to deploy successfully.

Scalable AI, Blockchain and Cyber Solution for the Warfighter

David Metcalf, UCF

AI, Blockchain, and Cybersecurity (ABC) advances are reshaping the enterprise solutions that support the warfighter. This session provides a survey to explore use cases under development at University of Central Florida’s Institute for Simulation and Training including the ARO sponsored Blockchain and Quantum Defense Simulator for multi-protocol prototyping, modeling, and testing, Army TRACRChain Blockchain for automated range data from TRACR2, and Navy Project AI Avenger analysis of AI media scrubbing tools. A review of design, standards, early results, and scalability opportunities and issues will be shared. Synergy with other projects and next steps in ABC solutions to meet emerging requirements for cross-warfighter solutions will be presented. Tangible examples include a digital twin prototype to combine operational readiness and trusted career-spanning data from recruit to retire and a quantum computing cyber awareness and AI Assurance simulation platform. Using platforms like digital twins, quantum-as-a-service, large language models-as-a-service, and advanced simulations allow Commands to explore specialized use case, protocols, standards, and scalability before committing vital resources – leveraging modeling, simulation and analysis techniques such as NSF ICorps and Hacking for Defense. Concluding remarks include discussion of methods of collaboration between military, industry, and academia to leverage public university research and other nonprofit entities.

Zero Trust to Empower Mission – Building the Enterprise Data Architecture for SAP/CAP

Josef Allen, USAF
Adam Gruber, Applied Insight
James Skelton, Applied Insight

Those defending our nation depend on access to accurate, timely information – and must manage large amounts of data from more sources now than at any other point in history. Disparate data sources, networks, and classification levels currently make it impossible for users in SAP and CAP environments to view data within a single standardized and normalized lens, limiting mission agility and increasing the time between data ingest and incorporation into command decisions.

To overcome these limitations, mission teams must currently develop custom tools and rely on manual processing of information to aggregate data and inform decisions. Feature gaps in pre-existing cloud capabilities within SAP environments further inhibit Guardians and other teams from efficiently leveraging cutting-edge technological capabilities to satisfy mission requirements, such as real-time data streaming, access to native cloud resources, and multi-cloud capabilities.

Providing holistic data processing in SAP environments presents three major challenges: data transfer across and between classification fabrics, data access governance, and multi-tenancy. Additionally, implementing a fully comprehensive Zero Trust Architecture is paramount.

This problem is complex, but with the right tools it is solvable.

To accelerate data sharing to mission teams in a Common SAP across classification fabrics and disparate networks, USSF built a highly scalable, multi-tenant, ATO’d environment – empowering program teams to migrate critical mission workloads to the cloud while maintaining logical separation of those workloads. Additionally, the USSF team designed and implemented a cutting-edge data management capability that enforces Zero Trust access to data assets leveraging a cloud-based architecture.

Federated Data Management at the Tactical Edge

Caden Bradbury, NetApp

AI models are only effective if they can be utilized in the most extreme tactic edge scenarios. (Think: in the back of a Humvee, on a Naval Ship, in a remotely operated drone, etc.) While the training of accurate models is vital, the biggest challenge in these edge environments is moving data and models to and from the tactical edge to core data center.

Models must be continuously improved to be used effectively. They must perform at the highest level possible for the DoD. This is especially true in life-or-death scenarios, like automated target acquisition models. To optimize models, new data must be continuously fed to the algorithm.

ExtraTerrestrial Networks: Exploring WAN in the Final Frontier

Wilbur Smith, Arista Networks
Lindsay Clarke, Arista Networks

In this presentation, we delve into the challenges and possible solutions when designing a unified, multi-domain network architecture that seamlessly integrates a diverse range of platforms: GEO & LEO satellites, airborne platforms, terrestrial networks, GovCloud transit, and trans-oceanic cables. This architecture not only ensures dynamic, encrypted, and secure multi-access networks, but also incorporates a self-healing fabric that can adapt to signal-denied environments while reducing operational load.

Complementing this vision, we will explore the paradigm shift from legacy network operating models towards a software-centric ‘modern operating model’. Here, configurations are procedurally generated by automation that incorporate variables from multiple discrete systems-of-record. We also simulate network changes in a virtual twin environment, deploy to the network upon completion, and generate comprehensive documentation of the change.

Leveraging AI/ML to Automate & Expedite Cyber Threat Hunting

Gentry Lane, ANOVA Intelligence

We’re looking for more DAOs to participate in a JFHQ-DODIN sponsored pilot. Nemesis is a novel computational software platform that addresses current capability gaps required to sustain deterrence in the cyber domain. By leveraging new discoveries AI/ML/DL, Nemesis functions as a global, interoperable, standardized cybersecurity software platform for early detection, rapid attribution, secure intelligence sharing, and global situational awareness of advanced persistent threat activity in DoDIN endpoints. Nemesis R&D was funded by DHS & DoE at Oak Ridge National Laboratory.

Hyatt Regency Portland at the Oregon Convention Center

375 NE Holladay Street
Portland, OR 97232
Hotel Phone: 971-222-1234
Hotel Website

Distance to the Oregon Convention Center – .1 miles
Distance to Light Rail – Adjacent

Per Diem Rate – $152 – Reserve Online Government
Industry Rate – $199 – Reserve Online Industry

Cut-Off date – Monday, November 27, 2023; if the block has not been filled already.

Hotel room rates are subject to applicable state and local taxes, currently 16%, in effect at the time of check-in.

Crowne Plaza Portland Downtown Convention Center

1441 NE 2nd Avenue
Portland, OR 97232
Hotel Phone: 503-233-2401
Hotel Website

Distance to the Oregon Convention Center – .5 miles
Distance to Light Rail – 4 Blocks

Per Diem Rate – $152 – Reserve Online
Industry Rate – $209 – Reserve Online

Cut-Off date – Sunday, November 26, 2023; if the block has not been filled already.