Let’s get one thing clear: the Democratic party’s presidential nominee, Vice President Kamala Harris, was never the border czar, despite her political opponents’ attempts to label her as such. If Harris has ever had a Biden administration czarship—not with an official title but with broad authority to coordinate and direct multiple agencies, organizations, and departments on a multi-faceted policy priority—it was in artificial intelligence (AI). Strangely, this doesn’t seem to have come up a lot in the 2024 presidential contest, despite the presence of AI everywhere else these days. In fact, this role doesn’t even merit a passing mention on the “Meet Vice President Kamala Harris” page of her website even as she prepares to formally become the party’s presidential candidate at the Democratic National Convention in Chicago.
AI might lack the political resonance of the border today, but it is time we reconsider its significance to the average voter. As Harris graduates from a vibes campaign to one with more substance, the vice president should put a spotlight on the AI in her record. When AI is recast as a sweeping change that could affect jobs, income equality, national security, and the rights of ordinary citizens, it is rather quickly transformed from esoterica to an everyday concern. The Trump-Vance campaign has received support from the likes of Elon Musk, Peter Thiel, and Marc Andreessen—all major Silicon Valley and AI influencers and investors—but it is Harris, not former President Donald Trump, who has actual fingerprints on AI policy. So, what has been Harris’s track record in this area? And where is the vice president likely to take AI policy if she wins the White House?
Harris’s role as AI czar may be the political season’s best-kept secret. But if one were to trace AI policy development in the world’s leading AI-producing nation, all signs point to Harris. Remarkably, AI policy development has been led by the White House rather than the U.S. Congress. In fact, Congress has done precious little, despite the growing need for AI guardrails, while the White House, with Harris as the seniormost public official involved, has helped frame and follow up on its October 2023 executive order on AI.
That order was designed to ensure the “safe, secure, and trustworthy development and use of AI.” In addition, Harris has made a broader commitment to “establishing a set of rules and norms for AI, with allies and partners, that reflect democratic values and interests, including transparency, privacy, accountability, and consumer protections.” Significantly for a technology disproportionately reliant on a handful of industry players, Harris suggested and has led the important first step of bringing these players together to commit to a set of AI practices and standards that advances three critical objectives: safety, security, and trust.
Given the disproportionate influence of the United States on AI used around the world, it is critical for the country to have its public position clarified in international fora. Harris has represented the United States in key international convenings and led the country’s global advocacy efforts on ensuring safe AI, such as at the AI Safety Summit in Bletchley Park, England. At the other end of the stakeholder spectrum, Harris has also met with the communities most directly affected by the wider adoption of the technology, including consumer protection groups and labor and civil rights leaders, to discuss protections against AI risks.
Harris’ contact with AI has another dimension, too. As the ultimate political unicorn—a woman of color, an underrated and parodied vice president in the Biden administration, and an overnight sensation as presumptive presidential nominee of the Democratic party after U.S. President Joe Biden stepped aside from the 2024 race—Harris’s narrative has been defined largely by others, whether it is through AI-assisted disinformation campaigns or viral memes. She has been personally targeted by deepfake videos of her supposedly making garbled statements, such as, “Today is today, and yesterday was today yesterday.” An AI-aided voice synthesis that led to a demeaning parody of her presidential campaign advertisement was reposted by Musk himself on X. Trump has also falsely claimed that the large crowds at Harris’s campaign rallies were AI generated. In other words, Harris can legitimately claim to have had AI weaponized against her personally.
Finally, Harris hails from the global capital of AI. As former attorney general and senator of California, she has been financially supported by many in the tech industry; more than 200 Silicon Valley investors have backed her run for the White House. One of her closest confidantes is her brother-in-law, Tony West, Uber’s chief legal officer (now on leave to work for the Harris campaign). It is legitimate to ask if she would be willing to confront the industry on difficult issues; at the same time, her closeness with industry leaders could help with greater government-industry collaboration.
What can we learn from Harris’s record as to what she would do in the presidency on this issue? As AI czar, Harris showed some clear patterns. For one, her primary focus has been promoting safety and addressing the risks of unregulated AI use, which can lead to bias or abuse. Second, the White House under her stewardship has accomplished a wide range of safety-, security-, and trust-enhancing actions since the issuance of the executive order—from AI testbeds and model evaluation tools developed at the Department of Energy to the Office of Management and Budget-issued government-wide policy on AI, the latter with safeguards to assess and monitor AI’s societal impact. There have been pilots at the departments of Defense and Homeland Security using AI to protect vital government software and a call to action from the Gender Policy Council and Office of Science and Technology Policy to combat AI-generated image-based sexual abuse. Harris has also been the seniormost official behind the release of a Blueprint for an AI Bill of Rights, outlining principles for the ethical design and use of AI.
Another of Harris’s initiatives has been aimed at promoting authenticity as concerns about AI-generated content skyrocket. This includes proposing international standards for tracing authenticity of government-produced digital content and identifying AI-generated or manipulated content, through digital signatures, watermarking, and labeling.
While Harris is not, any means, an expert on AI, and much work remains to get a full-throated AI policy in place, the numbers tell a tale of steady early accomplishment. A list of 100 action items following the executive order has been completed by various federal agencies on issues ranging from developing new technical guidelines for AI safety to evaluating misuse of dual-use foundation models and developing frameworks for managing generative AI risks. Harris has obtained voluntary commitments from 15 companies to ensure safe, secure, and transparent development of AI technology. 31 nations have joined the United States in endorsing a declaration establishing norms for responsible development, deployment, and use of military AI capabilities. And the U.S. government has won commitments of up to $200 million from 10 leading foundations to fund work around five pillars that cover issues from democracy and rights to improving transparency and accountability of AI.
Harris’s campaign rests on the idea of looking to the future and “not going back.” The Democratic National Convention in Chicago presents an opportunity for Harris to communicate more to the public about a key part of that future: AI’s economic and societal implications and her role in influencing them. Time is running out on conveying this issue’s importance, especially to the working class. While the impact of AI on different occupations is a matter of debate, some argue that, in the near-term, higher-income workers are more likely to benefit from productivity improvements due to AI and the share of income going to capital is likely to increase at the expense of the share that goes to labor. Both trends would contribute to an increase in income inequality.
As for the impact on jobs, there are different schools of thought. Some believe AI could help make many services, such as medical care, or currently elite job responsibilities, such as research, writing. graphics design and software coding more accessible to the middle class. Others see a plausible scenario of a hollowing out of specialized job functions. Policy and election promises need to show how a Harris administration would help steer toward the former outcome.
On the global stage, there are numerous existential risks associated with AI. Autonomous lethal weapons are a critical concern as multilateral agreements to ban such weapons have failed. Tensions with major AI-producing nations such as China are escalating, with no roadmap for getting to common ground as both the United States and China have declared their aspirations to become the world’s AI leader. A recent seven-hour meeting between top officials of the two countries in Geneva advertised as a dialogue on managing AI risks reportedly ended with no concrete agreements or follow-up meetings scheduled.
In parallel, the atmosphere has only become more tense with U.S. tariffs on Chinese imports and restrictions on the export of high-end chips to China. The Commerce Department is considering further restrictions on exporting proprietary AI models to China. Meanwhile, Beijing and Moscow are discussing a strategic partnership on various issues, including technology, while the Chinese embassy in Washington has accused the United States of “economic coercion and unilateral bullying.” If mishandled, these tensions can escalate.
Harris’s campaign can distinguish her candidacy with an acknowledgement of her track record and momentum on AI policy development. It must make the case for at least three sets of issues her administration would address. First: understanding AI’s impact on jobs and the resulting impact on economic inequality, and setting forth a plan to mitigate risks and protecting the most vulnerable. Second: developing a strategy for harnessing AI that addresses key kitchen-table concerns, such as accessible healthcare and education and skill-building. And third: crafting a vision for U.S. leadership in AI that advances responsible innovation, reduces geopolitical tensions, and preserves American national security interests.
Going from czar to president is unusual and comes with unusual challenges. Czars are usually not formally appointed as such—Harris was never officially designated AI czar despite the clear czar-like nature of her involvement—but can work to bring multiple parties together, often doing so outside public view. Then-Special Presidential Envoy for Climate John Kerry, for example, played a key role as de facto climate czarin increasing cooperation on that issue with China without much fanfare.
In other instances, and when they are brought in during an acute crisis, czars come with enormous expectations: The city of Boston awaits a rat czar, and residents want to see quick results. Czars do not have executive powers but have the respect of many, which is the calling card that allows them to convene parties with differing agendas. Presidents enjoy none of these luxuries. They own the problems they take on and they do so in public view.
There’s no escaping the reality that we are—and this election is being held—firmly in the age of AI. It is important that Harris’s team conveys the significance of AI to people’s lives and lets voters know how Harris would build on her unique track record. American voters have a choice to make for the nation’s next president this November, and on this one critical issue at least one of the candidates has a running start.
The post If Kamala Harris Was the Czar of Anything, It Would Be AI appeared first on Foreign Policy.
