In these uncertain times, we face the emergence of increasingly sophisticated generative AI models, or GenAI. These models promise to transform society, revolutionize various industries, and drive innovation across multiple sectors. In addition, this technological development potentially marks a significant step towards the creation of more powerful forms of AI in the future, including Artificial General Intelligence (AGI).1
However, these advancements come at a time of complex and unstable geopolitical conditions which bear resemblance to the era before World War I. During that period, industrial growth and an uncertain security environment, along with clashing ideology, led to a conflict unparalleled in its ferocity and destructiveness.2
Today’s technological breakthroughs, while filled with potential, also bring risks. The geopolitical environment should be a stark reminder of the complexities associated with massive shifts in power and technological capability. As we move towards the development of more advanced AI, we must navigate these complexities with caution, wisdom, and a keen awareness of history.
“This moment provides the United States government with a unique opportunity to lead with conviction as humanity enters a new era.”
The competition for power and influence in this era will require that the United States, alongside allies and partners, develop government-supported, public-private partnerships that enable continuous technology innovation and deployment. This moment provides the United States government with a unique opportunity to lead with conviction as humanity enters a new era.
Two Principles Should Guide U.S. Action
- Global Leadership in Generative AI Should Be a U.S. National Security Priority. As the world enters the Age of AI, America’s continued leadership in this field is a crucial national security imperative. Crucial because GenAI has the potential to revolutionize military capabilities, intelligence gathering and analysis, and cyber warfare, all of which means the United States and its allies and partners must maintain a competitive edge. The ability to develop and deploy advanced AI systems can enhance a nation’s strategic advantage, deter potential adversaries, and ensure national security interests are protected. A failure to prioritize AI leadership would leave the country vulnerable to technological dominance by rivals, compromising its security and diminishing its influence on the world stage.
- Winning the Platforms Competition is Critical for our National Prosperity. After decades of growth and expansion, the PRC is now facing significant economic headwinds exacerbated by mounting policy challenges. Still, Beijing is relentlessly pursuing its goal of becoming “the world’s primary center for science and high ground for innovation,” which it sees as key to shaping the global balance of power to its advantage.3 While its current GenAI efforts lag behind U.S. platforms, the PRC is among a very short list of nations capable of building and deploying frontier large language models (FLLMs), given its top down leadership structure, immense resources dedicated for this effort, and innovative but constrained talent pool. GenAI is opening a new front in the technology competition. With both the PRC and the United States offering competing visions of the future trajectory of geopolitics, we are entering a new era of uncertainty that could become highly destabilizing, with potentially grave consequences for humanity.
Two Objectives Will Define Success
- The United States, Together With Its Allies and Partners, Lead AI Innovation and Set the Rules of the Road. The U.S. government, in close collaboration with its allies and partners, must seize the transformative potential of GenAI and strengthen our collective competitiveness for this new era. This pivotal moment demands swift action to ensure economic prosperity, build military strength, and promote social well-being. Given AI’s global nature and its potential impact on societies worldwide, it is imperative for the United States to take the lead in establishing a framework of rules, norms, and ethical standards governing the development and deployment of AI. This community of like-minded nations must work together to counter bad actors – state and non-state actors both – who will invariably use GenAI in malicious and destructive ways. Through proactive engagement and strategic investments, we can harness the power of GenAI to shape a future where the United States and its partners thrive in a world characterized by stability, progress, and shared values.
- The United States, Together With its Allies and Partners, Work to Avoid a Destabilizing AI Arms Race. The potential military applications of AI, particularly to facilitate kinetic and cyberattacks, along with the relatively low barriers to entry of AI capabilities, could fuel an escalating global arms race with ever more powerful weapons available to a wide range of actors. To prevent such a catastrophic outcome, it is imperative for nations to engage in dialogue, cooperation, and the establishment of shared rules and norms around AI, and regulations and enforceable consequences addressing AI weapons. The United States should lead global engagement to promote transparency, foster trust, and encourage collaboration. Together, the international community can mitigate the risk of a destabilizing arms race, ensuring that AI technology is harnessed for the benefit of humanity rather than becoming a destructive and destabilizing force.
Three Moves Are Required to Achieve Our Objectives
- Set the Conditions. To sustain and advance U.S. leadership in GenAI, the U.S. government must help its domestic actors by leading through a positive vision, getting the basics right, and rallying allies and partners in making similar, complementary investments. While GenAI is one of the many applications of AI, it shares the same basic requirements needed for broader AI adoption and advantage: data, compute, people, and a thriving innovation ecosystem. The U.S. government should accelerate its efforts to make quality, large datasets more widely available to U.S. organizations, particularly academia who lack access but are on the forefront of research; increase compute access while maintaining technological leadership; make targeted investments in those areas where private sector capital might be lacking or insufficient; and provide greater support in developing the next generation of technical innovators.
- Shape the Terrain. To ensure continued U.S. leadership in this era of innovation power, the U.S. government must lead in convening a range of stakeholders to set responsible rules, both domestically and internationally. Because GenAI presents new challenges to privacy, security, and societal cohesion, the U.S. government must work closely with industry and international partners to shape its development and impact. By fostering cooperation, coordination, and joint research and development, the United States can shape the trajectory of AI development and promote a global ecosystem that prioritizes human welfare, privacy, and the rule of law.
- Accelerate Implementation. The U.S. government must proactively incorporate GenAI into its daily work or risk falling behind geopolitical competitors and failing to maximize value creation for its citizens. Each branch, department, and agency of the U.S. government should be experimenting with how to responsibly incorporate GenAI tools into their mission sets. The nature of government work will involve unique requirements and constraints, but this should not serve as an excuse for inaction. Generative AI has the capacity to enhance government abilities, from administrative tasks to critical military operations, thereby boosting our national competitiveness.
As more powerful AI models – particularly AGI – increasingly come into view, it is crucial for our country to seize this moment and lead. We should approach GenAI – and AI more broadly – not with irrational fears, but with the intellectual curiosity and rigor that is the hallmark of our innovative nation and that positioned the United States as the global technological leader throughout the 20th century and continues now into the 21st. We must not be paralyzed by a desire to eliminate every risk. We need to articulate a positive vision; responsibly harness the opportunities of GenAI; and, with our allies and partners, serve as a model as we chart a new path for human civilization.
Part I of this report explores the geopolitical implications of recent breakthroughs in GenAI. It presents the national security context, technological trajectory, and potential impacts of this disruptive new technology to inform effective, coordinated U.S. government actions.
Part II of this report includes a series of memorandums to senior U.S. government officials on the near-term implications of GenAI specific to their domains, along with policy recommendations for adapting to rapidly changing conditions. The topline takeaways from each memo are as follows:
Memorandums to the President and Congress
Memo to the President and Congress on Innovation Power for the GenAI Flywheel
GenAI, unique from the technological revolutions preceding it, is changing the very nature of innovation. The first nation to master this step change will unlock a new form of soft and hard power – innovation power – by which military capabilities, economic prosperity, and cultural influence can be forged. The U.S. government needs to come together to organize, drive, and fuel the U.S. innovation ecosystem with empowered, tech-focused institutions; audacious moonshots; and new resources to convert its current technological edge into long-term advantage for the era of GenAI.
Memo to the President and Congress on Governance of Generative AI
The governance landscape for GenAI is being shaped by two contrasting timelines: GenAI development is progressing rapidly, while AI governance is moving slowly. The need for AI governance is urgent because of the rapid societal uptake of GenAI and the imminent and pressing threats that GenAI presents, such as accelerating the spread of disinformation. For the sake of America’s own democratic future, and given the global ideological contest over governance models, we must present the world with a democratic model for GenAI governance worthy of emulation, one that reaps the benefits of GenAI while also protecting our societies from the worst potential harms.
Memo to the President and Congress on Building the Generative Economy
GenAI will have a transformative impact on the global economy. Nations that harness the potential of GenAI will increase productivity, a key economic measure that underpins economic growth, higher standards of living, and the ability to finance all other national priorities. Unlocking these benefits requires boosting competitiveness across the fundamental building blocks that underpin AI leadership: compute, data, and people. Compute is the engine that powers GenAI training and inference. Data serves as the fuel, but competitiveness hinges on digital infrastructure to move it from place to place, as well as a trusted regulatory framework that can ensure responsible use. Ultimately, people are the most important element for GenAI leadership. We must continue to attract the world’s top AI talent while unlocking opportunities for Americans to benefit from the Age of AI.
Memo to the President and Congress on Establishing U.S. Global Leadership in the Era of Generative Artificial Intelligence
GenAI is transforming the global environment in which America’s diplomats operate and the tools with which they work. Success in the Age of AI requires the United States, with the Department of State in the lead, to implement a paradigm of “Platforms Statecraft” to ensure that our nation and our allies and partners can work together to support and promote our global technology platforms so that they underpin advantage in GenAI and other strategic technology areas. Simultaneously, the Department of State should adopt new tools that capitalize on GenAI’s transformative power in executing its mission. Finally, the Department of State should lead an international effort to design and organize multiple layers of international regimes, institutions, and dialogues to mitigate GenAI’s high-consequence risks and support the broadest possible positive-sum applications that GenAI offers.
Memo to the President and Congress on the Adoption of Generative Artificial Intelligence for Military Purposes
America’s advances in generative artificial intelligence present the Department of Defense (DoD) with a crucial opportunity to accelerate two of its most significant transformations – preparing for the future character of conflict and strengthening our military overmatch against our rivals, especially the People’s Republic of China. The establishment of Task Force Lima on August 10, 2023, was a critical recognition of the importance of this moment by DoD. We propose that the Task Force prioritize four critical areas in integrating generative AI across DoD: enabling decisional advantage, enhancing operations, developing talent, and identifying new defensive measures.
Furthermore and mindful that advances in generative AI will continue to accelerate in the near-term, we also recommend that the Department: 1) create a Defense Experimentation Unit to provide for much-needed operational experimentation and iteration with AI models; 2) build an Automated Orchestration Platform – a generative AI-powered interface that can call up relevant tools and datasets, can decompose the queries into discrete tasks, and semi-autonomously or autonomously complete tasks; and 3) develop defense-tailored generative AI models, trained on specific military information, terms, and jargon.
Memo to the President and Congress on the Implications of Generative Artificial Intelligence for the U.S. Intelligence Community
Rapid advancements in GenAI, and AI more broadly, make it clear we are on the threshold of the next era of intelligence, one that will be defined by how well intelligence services leverage AI tools to collect, sift, and analyze global data flows to generate insight and deliver effects. The U.S. Intelligence Community (IC) should take immediate action to leverage these emerging capabilities to protect the nation and maintain our competitive advantage over the PRC.
- Sébastien Bubek, et al., Sparks of Artificial General Intelligence: Early Experiments with GPT-4, arXiv (2023).
- Tristan Bove, Henry Kissinger Says The U.S. And China Are In A ‘Classic Pre-World War I Situation’ That Could Lead To Conflict, But A.I. Makes This ‘Not A Normal Circumstance’, Yahoo Finance (2023).
- Ben Murphy, et al., Xi Jinping: ‘Strive to Become the World’s Primary Center for Science and High Ground for Innovation’, DigiChina (2021).