Early 20th century leaders didn’t know much about oil, electricity, or energy, while they ended the century with a clear understanding of the dependencies of their nations on reliable sources of power, the conversion and transportation to their citizens, the final consumers, users and voters.
The trajectory of the 20th century was significantly shaped by nations that strategically positioned themselves in the global energy landscape. Access to, and control over, energy resources became a defining feature of economic strength and international influence. As we navigate further into this century, a similar paradigm shift is occurring, this time driven by technology, with Artificial Intelligence (AI) at its epicentre. For national leaders, understanding the fundamental components of AI is no longer a niche requirement but a strategic imperative. The United States, for instance, stands at a critical juncture in AI development and deployment. While maintaining significant advantages, it faces mounting challenges that require a clear and informed approach.
Just as with previous industrial and technological revolutions, it is extremely complex and challenging, if not impossible, for any single nation to achieve preeminence across every facet of the AI ecosystem. However, wise leadership involves making informed decisions about where to focus national efforts and resources. A foundational understanding of the “AI Technology Stack” is paramount for such strategic calculus. This stack can be conceptualised as a series of interdependent layers, each presenting unique opportunities and challenges, which we review below.
The AI Technology Stack provides a framework for comprehending the various components that collectively enable AI capabilities. For esteemed leaders and diplomats, a clear grasp of these layers is beneficial for policy formulation, investment decisions, and international diplomacy.
Governance Layer: Overarching the entire stack is the governance layer. This encompasses the policies, guidelines, and ethical standards that direct the use of AI within an organisation or nation. It includes ensuring compliance with regulatory requirements and the prudent management of risks associated with AI deployment. Effective governance is crucial for fostering trust, ensuring accountability, and aligning AI development with national values and societal interests. i.e. NIST, GDPR, AI Act, or the AB 2013 from California.
Application Layer: This is where the functional capabilities of AI are delivered to end-users. It comprises user interfaces and Application Programming Interfaces (APIs) that allow individuals and systems to interact with AI-driven functionalities, such as chatbots, predictive analytics tools, or image recognition software. The design and deployment of applications must be user-centric and aligned with specific mission outcomes. i.e. Netflix, Salesforce, WhatsApp,… this is what most of us see and interact with day to day.
Security & Privacy Layer: As AI systems become more pervasive and handle increasingly sensitive information, the security and privacy layer is of critical importance. This involves the tools, processes, and practices designed to protect systems and data from unauthorised access, ensure data integrity, and uphold user privacy. For Federal agencies, it is critical to prioritise infrastructure and applications that best protect the freedoms of the populace and promote human flourishing, particularly measures that safeguard individual privacy and sensitive data. Robust access controls, efficient data deletion mechanisms, effective data minimisation practices, and comprehensive data lineage tracking are vital components. Imagine you tried to access your IRS account and suddenly found yourself logged in as someone else. i.e. Crowdstrike, Darktrace, Palo Alto Networks, Tanium, HiddenLayer.
Algorithm Layer: This layer pertains to the selection, training, and refinement of machine learning models and other algorithmic approaches that form the core intelligence of AI systems. It includes the tools and frameworks used to develop these models. The sophistication and appropriateness of algorithms determine the efficacy and reliability of AI applications, from predictive analytics to complex decision-support systems. Remember the impact the DeepSeek algorithm had on the entire AI field a year ago? i.e. OpenAI, Anthropic, Hugging Face, Gemini.
Data Layer: AI, particularly machine learning, is fundamentally data-driven. This layer involves the storage, management, and processing of the vast quantities of data that AI systems utilise and generate. Critical considerations at this stage include ensuring data integrity, security, and accessibility. For government applications, robust data governance is crucial to facilitate testing, evaluation, validation, and monitoring of AI systems. Investing in foundational data infrastructure is key to supporting operational precision in AI applications across all sectors. Imagine you’re a Michelin-star chef: the difference between arriving at your kitchen with all your tools and ingredients organised and cut, ready to cook, and arriving to find them hidden and disorganised. That’s what it feels like with data. i.e. Databricks, Scale AI, Appen, Snowflake, Collibra.
Infrastructure Layer: This is the bedrock of the AI ecosystem. It encompasses the underlying hardware and software resources essential to support AI applications. This includes tangible assets such as servers, advanced computing chips, network devices, and cloud services, as well as other computing resources. Without a robust and accessible infrastructure, the development and deployment of AI at scale are unfeasible. Strategic decisions regarding domestic production, international partnerships, and supply chain resilience for these resources are of paramount importance. i.e. NVIDIA, AMD, Intel, Microsoft, Google, AWS.
And guess what layer comes below to support the Infrastructure and data centres…that’s right, the energy to power that infrastructure. Everything is connected.
The global landscape is increasingly influenced by the rapid advancements in Artificial Intelligence. As with the strategic command of energy resources in previous eras, nations that comprehend and strategically navigate the complexities of the AI Technology Stack will be better positioned to secure their interests, foster economic prosperity, and enhance national security.
It is not expected that leaders become technical experts in every layer of this intricate stack. However, a commanding understanding of its components, their interdependencies, and their strategic implications is indispensable. Only by surrounding themselves with competent counsel and fostering a culture of informed decision-making can leaders effectively harness AI’s transformative potential. This will ensure that AI development and deployment align with national priorities and contribute to a future in which technology enhances human flourishing while maintaining stability in a world increasingly characterised by complex challenges.
Success in this endeavour will necessitate unprecedented cooperation between government, industry, and academia, underpinned by strong international partnerships.
