5 ways the US can drive the world’s AI agenda 

Much of the handwringing about AI over the past year has been about whether it will help humanity or invite some dystopian abyss. Less attention has been paid to what bad global actors can and will do with the technology (or already are doing), and how a Wild West approach to AI could endanger not just jobs but eventually national security. 

This moment is an opportunity for the U.S. to not just lead the world in AI private investment, innovation and development. We’re already doing that. The U.S. government must now begin the hard work of establishing a global AI regulatory framework — one that advances democratic values and sets the AI rules of the road for decades to come. 

Data is already playing a critical role in the race for AI supremacy. To develop ever-more sophisticated and accurate AI systems, companies need access to high-quality data at unimaginably high volumes. This is why regulating data and AI, and creating the proper guardrails, will be one of the most pressing issues in Washington over the next year. 

Part of the path toward doing this is taking on the threat of China more directly, and the foreign aid bill Congress just passed — with language forcing the sale of the Chinese-owned social media site, TikTok — is a good start. The main reason for ridding TikTok of its Chinese parent, Byte Dance, is concern that it could be sharing data from the site with the Chinese government. This has significant national security implications and drives home the importance of U.S. global leadership in AI, and AI governance. 

To maintain its lead technologically and advance its AI governance model globally, the U.S. needs to take several concrete steps: 

More R&D, education funding: For the U.S. to maintain its lead in AI development, the federal government and Congress need to increase R&D funding for AI to overcome China’s data advantage. China has prioritized investment in its AI talent pool,and the U.S. needs to do the same by increasing education funding and skills training for the AI future. For now, the U.S. leads the world in terms of total AI private investment — in 2022, the U.S. invested $47.4 billion — and China is second with $13.4 billion. However, across China, AI is being deployed much quicker than in the U.S., and in 2021, China installed more industrial robots than the rest of the world combined. The U.S. must close gaps with China where they exist. 

Get Privacy Regimen in Order: On global AI governance, the U.S. must finally pass a comprehensive federal privacy regime, key for both domestic regulation but also to build greater credibility for U.S. AI positions globally. 

Tighten Tech Restrictions: The United States must continue to escalate its restrictions on semiconductors and other key technologies going to China. The existing bans are incomplete, and chipmaking equipment and advanced chips continue to reach China, frequently through other countries. President Biden recently took an important step to prevent the large-scale transfer of Americans’ personal data to countries of concern, but such restrictions need to be extended to data sets and other AI-enabling elements.  

Intensify Global Leadership: The U.S. should escalate its participation in international standard-setting bodies to advance its standards, and counter China’s leadership in these organizations, which it uses to deploy its state-driven vision for AI. The U.S. must reengage in digital agreements, which will play an important role in AI governance.

Focus on Democracy: With elections this year in at least 64 countries, including the U.S., the increased sophistication of AI in facilitating misinformation and disinformation is of grave concern. The U.S. must lead the effort to regulate misinformation, minimizing false information spread by bad actors and protect democratic values. 

The Biden administration’s executive order on AI, released in October, is a good starting point. While lacking the force of law, it establishes industry guardrails, provides funding for R&D and talent development, and establishes a cohesive effort for AI deployment through federal agencies.  

The U.S. must use the broad federal government initiatives mandated by the executive order to set a global example in AI deployment and governance. This initial foray by the U.S. government will serve as a test case for assessing the impact of AI on the workforce, and for developing regulations that will both protect workers and train them for the new AI economy. 

And the data battles that may begin with TikTok will surely not end there. AI systems need trillions of bits of data, which are gathered from every publicly available source, including books, articles, websites and podcasts. U.S. companies such as Open AI and Anthropic are engaged in an existential race for data, even exploring creating synthetic AI-developed data to feed the artificial intelligence beast — something China is surely doing, too.  

With its massive surveillance state, China has an enormous advantage in terms of access to data, with upwards of 20 million cameras gathering facial and other types of data from its 1.4 billion citizens. Through its extensive Digital Silk Road initiative, China sells internet equipment throughout the Global South, providing cameras and surveillance equipment to replicate its state control model. Not only does this allow local authoritarian regimes to surveil their own populations, but China is also able to access that data from a wide range of populations. 

U.S. economic and national security depend on maintaining a lead in AI development and getting AI global governance right. The U.S. must use its budding domestic initiatives to lead the way on AI governance, ensuring that AI does not become a tool of authoritarians. Our future depends on it. 

Orit Frenkel, PhD, is co-founder and CEO of the American Leadership Initiative. She is a former senior executive at the General Electric Company and served as director for trade in high-technology products at the Office of the U.S. Trade Representative. Follow her on Twitter @OritFrenkel and @AmerLeadInt.