BY MARTHA LINCOLN AND MARTHA KENNEY

Earlier this spring, we published an essay in Academe about the California State University’s secretive and costly initiative to provide ChatGPT Edu, a general-purpose AI chatbot, to all CSU faculty, students, and staff. With many colleagues at Cal State, we have spent the past year organizing against the “AI-Empowered CSU.”

On May 19, we learned that the CSU administration will be renewing this controversial contract. Over the next three years, the cash-strapped CSU system will pay OpenAI a total of $39 million, or over $1 million a month, for systemwide access to its proprietary chatbot. We see the renewal as not only costly, but counterproductive to the institution’s mission. As a recent New York Times Magazine investigation suggests, the outcome of CSU’s AI initiative thus far has been “chaos” across our 22 campuses.

By re-upping this contract, leaders of the nation’s largest public university system are also openly conveying their disregard for labor. Last year our faculty union, the California Faculty Association, filed an unfair practice charge against CSU management for failing to meet and confer about the impacts of the AI initiative on our working conditions. For CSU to renew a contested contract before the matter has been adjudicated shows bad faith.

It is also troubling that CSU has discounted student and faculty concerns about the ethical issues posed by AI corporations and their products. These issues include the environmental costs of data centers, OpenAI’s contracts with the US Department of Defense, allegations about mental health harms caused by ChatGPT use, and an emerging image of Sam Altman, OpenAI’s CEO, as a singularly duplicitous and mercenary figure.

This decision also lands amid a heated national debate about the negative effects of AI on the future of work. In recent weeks, graduating classes nationwide have booed commencement speakers for repeating industry talking points about the transformative power of AI in the labor market.

Though CSU has acknowledged students’ concerns about AI, it has not meaningfully addressed them. Announcing the contract renewal in a newsletter for faculty earlier this month, CSU stated:

When Chancellor Garcia and the Board of Trustees chose to invest in AI, we understood the decision would draw attention and, at times, criticism. We moved forward with a clear purpose: to ensure our students are prepared for a rapidly evolving world and to keep student success at the center of our work.

In this blog post, we briefly reflect on how CSU is justifying its new contract with OpenAI in the face of criticism from students, staff, and faculty. We suggest that the administration’s statements mobilize a discourse that we call AI centrism—a framing that attempts to balance the costs and benefits of artificial intelligence use and arrive at a compromise.

While AI centrism often sounds reasonable, it both avoids engaging with the most serious critiques of the technology and encourages the normalization of AI use. And while this discourse appears more moderate than that of the AI industry itself, at CSU it has been used to justify the uncritical mass adoption of general-purpose chatbots.

AI centrism positions itself as a happy medium: a middle ground between the techno-optimism of AI enthusiasts and the techno-pessimism of detractors. AI centrists assume that splitting the difference between exalting AI and rejecting it means accepting the technology—because, as their reasoning often goes, “AI is here to stay.” At the same time, concerns about AI’s potential harms are supposedly satisfied by a vague commitment to “responsible” practices.

For example, in CSU’s frequently asked questions about the AI initiative, the administration claims that “higher education cannot ignore” the shifts engendered by the arrival of artificial intelligence, “but it must approach it thoughtfully and responsibly.” CSU also promises to “move forward with care and intention” in its AI ventures. Details about what constitutes responsible, ethical, or “thoughtful” use are not provided. There are also currently no CSU-wide policies on the use of ChatGPT Edu in research, learning, or teaching.

The AI-centrist framing of “responsible use” does not constitute a legitimate plan to protect the university and its students against the potential harms of this technology. Nor does it address the social and environmental impacts of creating and operating large language models. If faculty and students oppose the energy and water costs of data centers and OpenAI’s relationship with the defense department, refusing it is the only option. There is no middle ground.

The CSU often appeals to equity to justify the OpenAI deal. While we agree that CSU students should have equitable access to educational opportunities, we do not see this contract as a means to that end. ChatGPT Edu has not been designed, trained, or optimized for education and often provides answers that are unsubstantiated and incorrect. Because ChatGPT Edu is not an educational technology, it does not require equitable distribution on our campuses.

Where CSU positions its deal with OpenAI in centrist terms, as a means of “thoughtfully integrat[ing] AI into teaching, learning, research and operations,” we suggest it is best understood as an unregulated experiment that puts CSU’s working-class students at risk of receiving a second-tier education.

Faculty, staff, and students should be shaping the future of CSU—not just administrators and industry leaders. The systemwide introduction of AI on our campuses poses financial, environmental, pedagogical, and labor concerns that cannot be resolved with bland appeals to “transparency, ethical use and responsible regulation.” If our university system is to address the challenges posed by AI in ways that are truly ethical, shared governance is the only way forward.

Martha Lincoln is associate professor of medical and cultural anthropology at San Francisco State University. Her research addresses the cultural politics of public health, with recent projects addressing medical crowdfunding and online health influencing. Her email address is mlincoln@sfsu.edu. Martha Kenney, professor of women and gender studies at San Francisco State University, is a science and technology studies scholar who works at the intersection of feminist theory, speculative fiction, and contemporary technoscience. Her email address is mkenney@sfsu.edu.