Promoting the Use of Trustworthy Artificial Intelligence in the Federal Government
The Trump administration outlined a plan for federal executive agencies to design, develop, and use “trustworthy Artificial Intelligence” to assist in accomplishing their missions. Executive Order 13960, signed on December 3, 2020, recognized that Artificial Intelligence (AI) has improved the “operations, processes, and procedures” of agencies and businesses already. This order is a continuation of the push by the administration to maintain dominance in the AI industry worldwide.
This policy comes years after two other prominent countries in President Trump’s foreign policy purview have planned their course for dominating the industry. On July 20, 2017, the Chinese government released a strategy on how they planned to lead the AI industry by 2030. Then, just two months later Russia’s President, Vladimir Putin, announced “ becomes the leader in the field will rule the world” when discussing how important AI is for the future of Russia.
This order is not the first published by the Trump administration in the AI arena. It follows a February 2019 Executive Order that establishes the administration’s game plan to maintain the American lead in the AI industry by launching the American Artificial Intelligence Initiative.
To bring more professionals into this AI initiative and foster expertise among younger professionals, the General Services Administration (GSA) is instructed to “establish an AI track to attract experts from industry and academia to undertake a period of work at an agency” under the Presidential Innovation Fellows program.
The executive order sets rigid deadlines to implement the promotion of AI in these executive agencies and puts the Federal Chief Information Officers Council at the forefront of implementing this policy and ensuring deadlines are met. These deadlines include creating an inventory of the AI already being used within agencies and the creation of interagency dialogue to continue to promote the use of this technology.
The order uses the definition of AI developed in the National Defense Authorization Act (NDAA) of 2019 to pursue the agenda of promoting more AI technology within the executive branch. That definition states that AI is an “artificial system that performs tasks under varying and unpredictable circumstances without significant human oversight.” The Act continues to describe AI as any system that “solves tasks requiring human-like perception” and “designed to act like a human.” While there are many aspects of AI, two of the main components of the technology are automation and no need for oversight.
The executive order sets out nine principles to “foster public trust” that each executive agency must follow when designing, developing, and using AI. This notion of fostering public trust is repeated throughout the executive order and the subsequent press release from the White House.
In 2019, Blumberg Capital surveyed 1,000 adults over the age of 18 in the United States to understand consumers’ level of comfort with AI and their overall sentiment toward the technology. The survey found that 48% of consumers lack confidence in AI’s ability to keep their private information secure and are concerned with cybersecurity attacks.
However, another important aspect this survey brought up is the fear or apprehension Americans have that AI will replace them in their jobs. Forty-nine percent of respondents feel that AI has already replaced positions and people have lost jobs. In contrast, 90% of tech execs somewhat or strongly agree that AI will take over mundane tasks, leaving the more creative and critical thinking tasks to humans.
The executive order mentions numerous times the importance of creating trustworthy AI and developing this technology in a way that will increase American trust in government agencies. This AI initiative brings to the forefront the issue of whether Americans trust this technology in general and whether they trust it being used in the government. These statistics and this constant reminder of creating “trustworthy AI” beg the question of how AI will be implemented within the government. Will it be used to promote efficiency and taxpayer savings at the expense of the federal workforce? Or will it be developed in a manner that will increase security, public trust, and innovation within the United States?
While the Executive Orders signed by the Trump Administration over the past two years have devoted attention to the promotion of AI within the government, little has been said of what sectors within those agencies will actually be promoting this technology. In addition, little has been said as to what kind of work will be taken over by AI. The public reaction to this initiative will not be fully understood until well into the next administration's tenure.
