The National Security Commission on Artificial Intelligence (NSCAI) was established by the US Congress in Fiscal Year 2019 to “advance the development of artificial intelligence, machine learning, and associated technologies by the United States to comprehensively address the national security and defense needs of the United States.” A few months ago, the 15 member Commission, - led by former Google CEO Eric Schmidt and former deputy secretary of defense Robert Work, - released its interim report. The final report is due to be released in March, 2021. It remains to be seen how the Covid-19 pandemic will impact the final report’s recommendations and release date.
“The convergence of the artificial intelligence revolution and the reemergence of great power competition… threaten the United States’ role as the world’s engine of innovation and American military superiority,” said the interim report in its opening statements. “How the United States adopts AI will have profound ramifications for our immediate security, economic well-being, and position in the world… We are concerned that America’s role as the world’s leading innovator is threatened… We know strategic competitors are investing in research and application. It is only reasonable to conclude that AI-enabled capabilities could be used to threaten our critical infrastructure, amplify disinformation campaigns, and wage war.”
Like the Internet, - let’s remember that ARPANET was started in the late 1960s by the US Department of Defense, - AI is usable for military and civilian purposes. But, - also like the Internet, - AI is a general purpose technology that can be used for a wide variety of applications, many of them beneficial, but others potentially threatening.
What threats does AI pose? The Commission considered a number of potential threats to US security and interests, including:
- erosion of US military advantage by strategic competitors led by China and Russia;
- diffusion of AI-enabled technologies to rogue states, terrorists, criminals and other non-state actors;
- accelerated cyber attacks to both military targets and everyday systems, from critical infrastructure to cars and homes;
- cyber-enabled disinformation campaigns and other threats to our democracy that can be distributed on a massive scale;
- violation of individual privacy and increased risk of human rights abuses; and
- serious mistakes and unintended consequences if public and private sector institutions adopt powerful AI technologies before they’re ready.
In an interview at a 2019 MIT conference, former US Secretary of State Henry Kissinger was asked if we need AI control agreements with Russia, China and other nations similar to the arms control agreements that he spent so much time negotiating during the Cold War. Dr. Kissinger replied that for arms control to be effective, the two sides needed to share information and agree to inspections. But such mechanisms don’t apply to AI because the transparency that was essential for arms control would be very hard to establish for AI threats. In addition, the variety and speed of cyber attacks make it much harder to develop adequate control systems.
AI has the potential to significantly advance national security. The report recommends that national security agencies adopt AI-enabled systems in key areas where AI can help them better understand and execute their missions. These include:
How We Defend America. AI-enabled systems can help protect our borders, detect and combat cyber attacks and safeguard critical infrastructures. In addition, AI systems can provide real-time risk assessment to first responders and help law enforcement anticipate threats and emergencies.
How Intelligence Agencies Make Sense of the World. AI algorithms can sift through vast amounts of data to detect and analyze potential threats, providing faster and more precise situational awareness and higher quality decision-making.
How We Fight. AI-enabled machines, systems, and weapons can help military leaders understand and quickly develop a coherent operating picture of the battlespace, make complex decisions faster, put fewer U.S. service members at risk, protect innocent lives and reduce collateral damage.
Fundamental lines of effort. While the Commission’s full analysis and recommendations have to wait until the final report is released next year, the interim report identifies five major lines of effort that are necessary to preserve US advantages:
Invest in AI Research and Development. “Federal R&D funding for AI has not kept pace with the revolutionary potential it holds or with aggressive investments by competitors… The U.S. government must identify, prioritize, coordinate and urgently implement national security-focused AI R&D investments.”
Apply AI to National Security Missions. Strategic competitors are catching up with the US technologically by investing in AI and other advanced capabilities, threatening to erode America’s military advantages.
Train and Recruit AI Talent. “Our defense and intelligence agencies need access to more people with AI skills and expertise, both in-house and outside of government.” The Commission noted that the American AI talent pool depends heavily on attracting and retaining top international students and workers because US universities aren’t able to meet the demand for student interest in AI at the undergraduate and graduate levels.
Protect and Build Upon U.S. Technology Advantages. Given the very nature of AI, protecting those technologies used for national security is very difficult for a variety of reasons, including: AI research has been largely decentralized and industry-driven; most AI research is open and publicly available; and America’s research universities welcome top minds from around the world. The US and China have close AI linkages including exchanges of people, research, and funding, but, at the same time, “China takes advantage of the openness of U.S. society in numerous ways - some legal, some not, - to transfer AI know-how.”
Marshal Global AI Cooperation. “The United States must enhance its competitiveness in AI by establishing a network of partners dedicated to AI data sharing, R&D coordination, capacity building, and talent exchanges.”
Basic principles. Finally, given the diverse views on the topic and the gravity of the challenge, the Commission developed seven consensus principles to help guide their work, recommendations, and national discussions:
- Global leadership in AI is a national security priority and requires a robust basic research environment;
- AI adoption for national security is an urgent imperative to protect the American people and US interests and to shape international norms for using AI;
- Government needs the assistance of industry and academia to maximize the promise of AI and minimize its national security risks;
- The US must cultivate homegrown AI talent and continue to attract the world’s best minds;
- Actions to ensure AI leadership must preserve the principles of free inquiry, free enterprise, and the free flow of ideas;
- National security officials should collaborate with the AI development and ethics communities to achieve safe, robust, and reliable AI systems; and
- Any use of AI must have American values at its core, including the rule of law.
Comments