Is Biden’s Executive Order on AI Too Hands-Off for Higher Ed?

On October 30, 2023, the White House released an executive order and accompanying fact sheet outlining the Biden administration’s plans for the implementation and development of “safe, secure, and trustworthy” AI in the arenas of health, diplomacy, equity and education. The order sparked discussion about how higher education would be impacted and supported through this transition, as well as whether the federal administration should be doing more for educators.

AI Developments Before the Order

Many institutions and businesses were using generative artificial intelligence well before the 2022 release of ChatGPt, which brought the technology to the mainstream. 

“At my place of instruction and in the majority of colleges and universities across the United States, ongoing initiatives are focused on comprehending the impact of generative AI within educational settings,” said Tony Kashani, professor of education in the Doctor of Education program at Antioch University and interdisciplinary scholar who researches the intersection of humanities and the digital age. “Subsequently, efforts are being made to develop policies that can effectively address the ethical and pedagogical implications of this technology.”

University of Texas

The University of Texas at Austin has been active in the generative AI space for more than four years. 

“Our initial guidance to faculty was focused on teaching about the tools and encouraging faculty to find ways to incorporate the tools into classes to teach topics that might be difficult to teach otherwise,” said Art Markman, vice provost for academic affairs at the University of Texas at Austin. “We have run several webinars for faculty to talk about how AI may support innovations in education. This approach is consistent with the executive order.” 

The webinars, which ran as a series in 2021, covered topics such as defining artificial intelligence and machine learning, how to apply AI and ethical questions surrounding AI outside the classroom. 

The University of Texas at Austin is also home to Good Systems, an AI-centric interdisciplinary consortium of experts in communications, community planning, engineering, robotics and other fields. According to Markman, the program launched in 2019 and has since invested in projects that seek to curb disinformation, plan for the rise of smart cities and prepare for human and robot co-working situations.

University of Michigan

Similarly, the University of Michigan was already incorporating AI and its use into its research and educational missions before the executive order.

“The university empaneled a committee to look into the many opportunities and challenges AI presents to higher education and make recommendations on how to adapt its policies and processes to take advantage of the many opportunities and mitigate the challenges, and above all, on how to support our faculty,” said Mingyan Liu, associate dean for academic affairs at the University of Michigan. 

In 2023, the committee released a report detailing its findings since its empaneling. Key findings included the potential for generative AI to “transform teaching outcomes by creating customizable learning pathways” and “aligning instructor objectives with individualized student needs.” Additionally, generative AI can “support enhanced administrative productivity and service quality […]. For example, career planning and support services may leverage various GenAI functions for career development.” 

The committee also identified several risks, including AI’s potential to “[create] fiction and [represent] it as truth,” and the “social weaknesses” AI systems may create through the use of deep fakes or AI companions that give “dangerous advice to children.”

A list of regulatory suggestions assembled by the committee designed to maximize benefits and minimize harm included accurate documentation of all uses of generative AI, whether on the part of teachers or students. In addition, the university administration should immediately develop plans for long-term support of faculty as they grapple with the inevitable implementation of AI in the classroom, as well as the creation of AI certification programs for instructors, among other measures. 

“In the realm of academia, particularly higher education, responses to innovations tend to be sluggish,” said Antioch’s Kashani. “This is even more pronounced when grappling with generative AI, which operates at the speed of light. Therefore, the responsibility lies with individual professors and technologically adept administrators to familiarize themselves with this technology and forge partnerships with AI companies. This collaboration is essential to ensure the establishment of an ethical framework for AI within higher education.”

Specifics of the Executive Order

Although the executive order’s accompanying fact sheet mentions education only within a single bullet point, the order itself provides a few specifics regarding the Biden administration’s goals for AI in higher education. Section 2(e) promises protections in many fields including education, without stating explicitly what those protections will entail. 

Meanwhile, Section 5(c) states that, within 180 days of the order’s issuing, the Secretary of State will “establish… a program to identify and attract top talent in AI and other critical and emerging technologies at universities, research institutions, and the private sector overseas.”

Section 6(b) adds that the government will, within 180 days of the order, “in consultation with the Secretary of Commerce and the Secretary of Education, strengthen and expand education and training opportunities that provide individuals pathways to occupations related to AI.”

Perhaps the largest impact is in Section 8(d), which promises the administration will aid in “the responsible development and deployment of AI in the education sector,” by developing resources, policies and guidance regarding AI. These resources will “address safe, responsible, and nondiscriminatory uses of AI in education, including the impact AI systems have on vulnerable and underserved communities.”

The section includes instructions for “the development of an ‘AI toolkit’ for education leaders implementing recommendations from the Department of Education’s AI and the Future of Teaching and Learning report. Such measures will address “appropriate human review of AI decisions, designing AI systems to enhance trust and safety and align with privacy related laws and regulations in the educational context.” 

Although the contents of the toolkit are currently unknown, Michigan’s Liu said, “U-M hopes to lead in appropriate development of the toolkit.”

Potential Impact of the Executive Order 

The lack of specificity in the propositions laid out in the order has left some skeptical of its ability to impact change or direction. 

“This executive order features ambitious rhetoric reminiscent of a campaign platform, brimming with grand promises of equity and justice for the American people,” said Kashani. “However, it lacks substantial language advocating for regulatory measures that could become enforceable laws.”

The order is specific in its fact sheet. There is explicit mention of “AI’s potential to transform education by creating resources to support educators deploying AI-enabled educational tools, such as personalized tutoring.”

“I think that if the executive order had been much more detailed than that, it would have incorporated suggestions that may not be the best approach to incorporating AI into education,” Markman said. “Instead, the shorter statement reinforces the importance of AI while still respecting the expertise of educators to focus on development of tools and curricula.”

According to Markman, AI is likely to be more effective as a tool in higher education to teach particular concepts.

“For example, in writing instruction, we are developing templates for teaching skills like audience design, revision, and outlining by incorporating AI into the learning process rather than assuming that AI systems will be effective at supplementing instruction,” he said. 

On the question of the executive order’s impact across higher ed, others were optimistic.  

“One of the stated objectives of this EO is to identify and attract top talent in AI and other critical and emerging technologies,” said Liu. “This could have a significant impact on higher ed, potentially increasing the number of international students and researchers in these fields.” 

Liu hopes that, as different aspects of the EO get implemented over the next few months, there will be additional research funding (or reallocation of existing research funding) in the directions central to this EO: safety, security, and trustworthiness of AI. 

“Our hope is that this executive order stimulates additional research related to our Good Systems program, as well as funding for the development of education resources to incorporate AI into education,” said UT Austin’s Markman. “Higher education has begun to move in the direction of creating Open Education Resources that can be used by all without cost. Open Education Resources will ensure that the benefits of AI systems are available to all.”

Moving Forward with AI

“Given how rapidly the field of AI is moving […] higher ed (i.e., its many individual institutions) has largely been left to define its own AI strategy and policy,” said Liu. “ This EO is a good start by the government to outline what it hopes to accomplish around the safety, security and trustworthiness of AI and will allow higher ed to better align its own strategy with the national plan.”

On the other hand, higher ed may continue to struggle to enact AI policy until the administration changes, according to Kashani.

The White House AI council, which was established alongside the release of the executive order, includes the Secretary of Education (or a designee thereof) and at least 27 other members. However, the order does not state how the council will “coordinate the activities of agencies across the federal government to ensure the effective formulation, development, communication, industry encouragement related to, and timely implementation of AI-related policies.” 

“We need individuals who possess genuine and democratic education, not just Ivy League social connections and questionable law school degrees,” he said. “Once we have lawmakers of this caliber collaborating with educators, ethicists and well-intentioned AI developers, higher education will receive the substantial support it needs.”