The for-profit, high-stakes Silicon Valley takeover of educational technology and AI innovation has squarely targeted our already gutted higher education institutions. The AI industry promises outsourcing for educators and staff, tutors for students, agents for administrators, and triage for institutions facing funding cuts. But the promises belie the more important reasons the industry is focused on higher education.
Teachers and learners generate rich data streams as they use educational technology systems. They also contribute valuable and often self-authored content and educational materials, curricula, texts, and data to these systems. With the proliferation of data mining, learning analytics and algorithmic diagnostic tools, and especially the wide-scale embedding of AI into legacy technologies, the procurement and implementation arrangements for ed-tech systems among institutions, vendors, and community members have become more complex and opaque. Time and time again, faculty are cut out of institutional and vendor decision-making regarding technology. Ed-tech companies see higher education as a cash cow. Once they are in the door with a vendor contract, their products are rarely removed, even when their contracts are expensive, their technologies unhelpful, and their data practices horrendous.
While ed-tech companies promise improved and personalized learning, reduced workload, and shared knowledge transfer, in our conversations with educators, many report that these outcomes do not materialize. The data and content fed into systems are used to train those systems. There is no way to know the extent of the intellectual property extraction and commodification that is occurring, nor how rights are being violated downstream by ed-tech and AI companies. Vague and ineffectual institutional policy structures in higher education allow administrators to enter vendor contracts that permit these companies to violate the rights of students as learners and educators as workers.
The tech industry’s overreach highlights the urgent need to “bring the fragments together” by building coalitions of unions, faculty senates, interested individuals, political movements, community organizations, and others interested in the fight against the corporate takeover of our society. Higher education is a promising arena for this broader fight.
To that end, the national AAUP’s ad hoc Committee on Artificial Intelligence and Academic Professions, on which the authors of this article serve, surveyed AAUP members about AI. Respondents reported that technology is subject to little or no policy oversight at most institutions, and even when faculty technology governance committees do exist, these bodies are often composed of hand-picked faculty members and staff who largely rubber-stamp administrative initiatives. To confront this problem, the committee recommended in its July 2025 report, Artificial Intelligence and Academic Professions, that academic workers, through their unions or through faculty senates, establish independent committees or governance bodies that can hold administrators accountable for their decision-making and their failures to serve the educational mission. These bodies should be composed of faculty members of all ranks, students, and staff and should have power to oversee, negotiate, and even refuse the procurement and use of ed-tech at their institutions. The ad hoc committee’s report laid out research-based principles to guide the oversight body’s work, specifying that it should
- have access to all parts of the procurement and deployment process;
- have veto power over ed-tech procurement and deployment;
- perform ongoing evaluations of ed-tech data flows and uses at the university and vendor levels;
- have funds allocated for these evaluations;
- have meaningful levers of enforcement (for example, the ability to rescind a vendor agreement when an ed-tech system is found to be harmful or not helpful);
- have the ability to suggest new policy around ed-tech;
- have mechanisms through which faculty, staff, and student members can hold administrators accountable for protecting faculty, staff, and student data; and
- act as a liaison with the broader higher education community.
These guidelines establish the basis for ongoing involvement of campus community members in decision-making around technology.
Selecting Model CBA Language
Existing collective bargaining agreement (CBA) language can in some cases be adapted to support robust technology governance structures. The CBAs discussed in this article come from three sources.
The first source is the “Negotiating Tech” inventory developed by Lisa Kresge, a member of the AAUP’s AI committee who is affiliated with the UC Berkeley Labor Center. The inventory contains 175 CBA documents with technology-related provisions, drawn from a review of over 500 contracts. Contracts included in the inventory were identified through keyword searches and manual review. Each provision was coded using a structured framework capturing four dimensions: bargaining strategy, technology type and function, worker and union rights, and employer obligations. The inventory spans diverse sectors—including health care, industrial work, and both white- and blue-collar occupations—and illustrates how unions negotiate across the technology life cycle, from planning and adoption to implementation, use, and governance. Although not exhaustive, the inventory provides a detailed picture of union engagement with digital workplace technologies and offers examples that higher education workers can adapt and adopt.
The second source is the “Resource Guide for Addressing AI in Higher Education” developed by the AAUP’s AI committee, which draws on “Negotiating Tech” and on bargaining and resolution language provided by the AFT. The guide is organized by the recommendations developed from the AAUP member survey and subsequent member engagement. It includes categorical frameworks and selected examples of contract language that AAUP members can bring to their local unions.
Finally, the third source is contract language, resolution language, and items used outside of contracts that AAUP members shared with the AI committee during ongoing organizing efforts around AI, bargaining consultation with locals, and the like.
In what follows, we discuss selected bargaining and resolution language related to three key topics addressed in the AI committee’s report: (1) independent governance bodies focused on technology, (2) intellectual property guidelines, and (3) job security and working conditions. While these are not the only topics covered in the report, they are the ones we have found to be of integral importance to higher education workers for reporting here. The examples discussed here serve as a pragmatic framework for faculty seeking to develop strategies that strengthen oversight of educational technologies. Collective bargaining chapters may use these examples as templates or “wish lists” to be refined by legal counsel and negotiated by local bargaining teams; advocacy chapters might adapt them into principles or guidelines for institutional practices or policymaking. All higher education workers can engage colleagues around these principles and guidelines to develop shared commitments regarding technology adoption and deployment. Such practices help higher education workers reclaim influence over the technological infrastructure that shapes teaching, research, and service, ensuring that it supports the right to learn and teach freely and protects the autonomy of educators and students.
Technology Governance Bodies
Higher education unions have negotiated access, authority, funding, and accountability mechanisms related to workplace technologies. Although many existing CBA policies do not specifically address AI, they apply broadly to data-intensive systems. Because AI is inherently data-intensive and affects working conditions, these provisions can be advanced as mandatory subjects of bargaining in higher education.
The CBA in place at Wayne State University stipulates that a committee “shall be established to deal with matters related to online education.” It goes on to state,
The committee shall make recommendations on the following matters:
- Rights of faculty members assigned to teach online courses, including rights to select course materials, methods of delivery, methods of interaction (including both asynchronous and synchronous methods), methods of examination, and appropriate virtual office hours.
- Intellectual-property rights of faculty members with respect to recordings of lectures, lab tests, demonstrations, class discussions and similar electronically recorded or accessible materials.
- Establishment of a fund for Online Course Training and Development under the Associate Vice President for Educational Outreach in an amount sufficient to support quality online instruction, along with principles for awarding grants from the fund to assist bargaining-unit members in developing new online courses or modifying existing on-campus courses for suitability for online instruction.
- Guidelines for the determination of workload, including appropriate class size, for those assigned to create or teach a “first time” online course.
- Guidelines for online-course preparation and teaching for bargaining-unit members on the tenure track that take into account the research and time required for the development of online courses.
- Guidelines for University provision of hardware, software, and technical support to bargaining-unit members who teach online to ensure adequate delivery or material, curriculum development, course planning, and student interaction.
While this example focuses on online learning, it still provides a model that could be adapted for all considerations around technology, from learning management systems that fall more directly under the auspices of what is detailed above to contracts with AI or AI-invested companies and AI tools that may be incorporated in the learning management system. Critically, the CBA provides a mechanism for oversight of technology, even if the language needs to be updated for the current moment.
A second example comes from Miami University in Ohio, where the faculty union negotiated the following memorandum of understanding on artificial intelligence:
Miami University (the “University”) and the Faculty Alliance of Miami, AAUP-AFT (the “Union”) enter into this Memorandum of Understanding (“MOU”) intending to be legally bound hereby.
Consistent with the collective bargaining agreement, the parties shall meet and discuss a process for reviewing the impact of Artificial Intelligence (“AI”) on bargaining unit faculty members and the University at its Labor Management meetings.
If concerns arise regarding the use of AI at the University, the parties agree to discuss those concerns in the Labor-Management meetings. As part of that discussion, the parties may also propose guidelines regarding the use of any specific program or tool that uses AI for consideration by the University.
Violations of this MOU shall be subject to Article 21 [grievance and arbitration].
This memorandum of understanding is an incredible win for the Faculty Alliance of Miami’s first contract. It frames AI as a mandatory subject of bargaining and sets forth a mechanism for addressing concerns about its use.
Higher education workers everywhere should be pressing their institutions on what corporate AI technologies are supposed to do. If a technology intervenes in teaching or learning, faculty unions can argue that it pertains to their work and is thus a mandatory subject of bargaining. Where faculty do not have collective bargaining rights, they should still have a role in decision-making through shared governance—especially when the technological interventions involve areas of the faculty’s “primary responsibility,” defined in the Statement on Government of Colleges and Universities as including “curriculum, subject matter and methods of instruction, research, faculty status, and those aspects of student life which relate to the educational process.”
The examples from Wayne State and Miami University offer a foundation for constructing committees with a role in AI governance and serve as precedents for unions and faculty senates seeking to negotiate stronger roles in technology decision-making. Faculty members may use these examples to craft contract demands, guide negotiations, support enforcement, or influence institutional policy development even in nonunionized settings. As educational technologies become increasingly intertwined with working and learning conditions, such frameworks are essential for ensuring oversight and protecting the rights and interests of higher education communities—and indeed, the mission of higher education more broadly.
Intellectual Property Protections
A key challenge around ed-tech governance concerns the intellectual property rights of faculty, who originally develop curricula and course content; staff, who produce a variety of work outputs; and students, who produce significant amounts of content in completing assignments, projects, and exams. Legal frameworks such as the Family Educational Rights Protection Act (FERPA) and the Child Online Privacy and Protection Act apply to the governance of student data, while copyright, licensing, and fair-use considerations complicate ownership of educational content within ed-tech platforms. Vendors’ intellectual property in platform infrastructure further complicates governance dynamics.
Although some institutions may formally recognize faculty ownership of pedagogical materials, enforcement is often at odds with FERPA and internal policies. In practice, institutions often fail to uphold these protections, to provide transparency about vendor agreements, or to prevent unauthorized reuse of copyrighted instructional materials. For example, course-shell replication policies within and across academic units remain inconsistent, and faculty and student work is frequently reused by institutions without permission or attribution.
An independent technology governance body run through either an academic senate or a faculty union would be more effective than the current technology governance bodies that are run by institutional administrators, in which critique of technology is shallow (if it exists at all) and any pushback against administrators’ decisions is minimal (if it even happens).
The following language from unions at the State University of New York and Rutgers University are good starting points, but local union leaders have stressed that stronger enforcement mechanisms are required. Both state university unions are bargaining over AI and intellectual property protection in 2026. United University Professions, the union representing faculty and professional staff in SUNY, states in its explanation of university policies,
Under the current SUNY copyright policy, faculty retain ownership of works produced in the scope of employment, including works produced for online education unless there is a written agreement between the University and the faculty member to the contrary. Putting it more specifically, SUNY and faculty may contract for ‘work-for-hire,’ authorize the work in advance by written agreement, and determine in the contract who the owner shall be. . . . In the absence of a written work-for-hire agreement, copyright ownership vests in the faculty.
Rutgers AAUP-AFT’s explanation of university policy states,
Course materials shall be protected in accordance with the existing Rutgers Copyright policy (50.3.7, 2007) which states that ‘faculty [have] rights to retain copyright ownership to the scholarly and artistic works they create. . . . Students typically will own the copyright to works created as a requirement of their coursework, degree or certificate program.’ Third party vendors shall not have any ownership of course materials in a learning management system.
On the one hand, it is clearly necessary to expand these provisions that pertain specifically to constituent data. On the other hand, building out more effective shared governance and enforcement of existing university policies through a union-led committee, as suggested above, or through the faculty senate—and pushing for more specific intellectual property protections—would be positive steps toward meaningful accountability and implementation.
Job Protections
People with whom the AAUP’s AI committee has spoken have expressed significant concern about their jobs and the conditions of their work. Some imagine their course content being replicated and reused by the university. Others see areas like student counseling—an important contact zone between students and the university—becoming hotbeds for AI’s replacement of human labor. The work conducted by course notetakers for students with accommodations, which used to be done by work-study students or graduate students, is already being farmed out to AI. More nefariously, consulting applications like Gray Decision Intelligence are being marketed to institutions to help them cut costs, which results in job and program eliminations, admissions and hiring freezes, and disproportionate negative impacts in arts and humanities departments. Finally, the AI glut also contributes to the increase in faculty workload. One faculty member noted in an interview that the “administration [attempts to] push faculty to incorporate AI into student writing assignments, following the argument that we cannot prevent students from using AI irresponsibly, so we must teach them to use it responsibly. . . . I am not paid for this extra work, and the support I receive for this is laughable.”
Existing contract language that protects jobs from other forms of technological disruption can be adapted to address AI. The Waterville Teachers Association, a K–12 union, won the following provisions promoting job security and maintaining working conditions in its 2025 CBA:
A. The parties acknowledge and confirm that participation in the Distance Learning Program shall not be used by the District to argue that the Association may have waived any rights that may exist to the exclusivity of bargaining unit work. The parties agree that the Distance Learning Program, in whole or in part, involves bargaining unit work in the provision of educational services to the children of the district.
B. No member of the bargaining unit on effective date of this agreement in a tenure area shall be subject to a reduction in force, in whole or in part, as a result of the district sending/receiving courses in that tenure area through a Distance Learning Program.
This language does not explicitly address the technology involved in distance learning, only the working conditions related to distance learning, which are inevitably facilitated through technology. An expansion of this language would be necessary for it to pertain to technology more generally.
Other contracts explicitly refer to technology, and even AI, as we have seen. For example, Contract Faculty United at New York University, a union affiliated with the United Auto Workers that represents full-time, non-tenure-track faculty, has developed article items that directly pertain to generative AI job replacement. The union shared with us the following proposed contract language, which is still under negotiation:
The Employer may not assign GAI [generative AI] or any work(s) generated by GAI to teach a course, course component, workshop, or tutorial; perform advising work; or serve as an instructor for any other teaching activity.
The Employer may not use GAI to produce a course description, syllabus, or other instructional material.
The Employer may not use any materials created by a Contract Faculty Member or any materials generated during teaching activities involving a Contract Faculty Member to train GAI or any work(s) generated by GAI unless the Contract Faculty Member provides explicit, written permission.
This proposed language would go a long way toward mitigating job losses related to generative AI and grant faculty more control over the use of technology. However, the NYU administration is refusing to negotiate this item, illustrating again the importance of pushing for technology as a mandatory subject of bargaining. Any time a university requires the use of a technology, unions should argue that it is a mandatory subject. At the same time, this example highlights how bargaining is not only about language; it is also a negotiation with administrators who view it as their job to run the university like a business. Unions must organize around demands for more power at the table. The NYU union is doing an admirable job in this regard. Op-eds, petitions, cultivation of political allies, structure tests, and demonstrations build power, which is crucial for winning demands.
Bringing the Fragments Together
Asserting faculty authority over technological infrastructure is essential for reclaiming the educational mission of higher education, which is the basis for an informed democracy. Technology is just one area in which the need for meaningful shared governance exists. But it is one with relatively few ready-made examples from which to draw. We hope that the sample language and suggestions provided here will aid in the development of better models.
These CBA and related examples illustrate how unions in higher education and in adjacent sectors address AI, ed-tech platforms, and other technologies (such as surveillance tech) by expanding shared decision-making and including workers and community members in oversight. They can be adapted for higher education bargaining units that seek greater influence over ed-tech contracts, decision-making, deployment, and review.
Because bargaining contexts vary, the examples collected here should not be adopted verbatim; rather, they should inform the development of local proposals crafted in consultation with legal professionals and negotiated within the adversarial dynamics of collective bargaining. Overall, the examples demonstrate the need to
- entrench rights around technology that already are granted in some partial way by enshrining them in CBAs or in other binding legal language;
- update language in CBA articles addressing various aspects of faculty work so that they explicitly apply to technology and AI and address such issues as intellectual property, data rights, and nonuse rights that ed-tech implementation raises;
- expand rights that already exist within CBAs; and
- enforce CBA articles that do exist and can apply to the current technological landscape and improve oversight and enforcement mechanisms in CBA articles that have yet to be incorporated in the contract.
National committees on AI, such as the AAUP’s ad hoc committee, as well as local offshoots can support these bargaining and enforcement efforts by developing frameworks that articulate rights over data and knowledge products. Power-analytic approaches highlight the need to privilege the perspectives and recommendations of those who participate in teaching and learning—the faculty members and students at the core of the mission of higher education—in all decision-making, especially decision-making around technology.
As debates over the regulation of ed-tech and AI continue—including issues of training-data boundaries, intellectual property, and transparency and openness—academic labor organizations are positioned to play a critical role. Because educational technologies directly affect academic freedom, faculty autonomy, and data rights, higher education workers and students are key constituents in these policy arenas. In both K–12 and higher education, the stakes are high. Many of the leaders of legacy ed-tech and newer AI companies have said that they want to see the end of teaching as we know it.
Beyond CBAs, there are many ways to intervene in the concerns discussed above. These include more grounded organizing activities that can and should happen both alongside and outside of collective bargaining agreements. Where there is no union contract, policies can be incorporated into the faculty handbook. Further, power does not flow through perfect contract or resolutions language. It comes from being organized and strategic about demonstrating and enforcing power that undergirds the language.
There are also other ways to consider moving forward. The AAUP’s AI committee is collecting and sharing resources for extracontractual organizing. Faculty governance can be exercised locally simply by working with colleagues in a department or smaller academic unit to develop a sign-on statement committing not to use corporate AI in specific ways. Collective statements and principles of this nature can be very powerful. Petitions for universities to walk back contracts with harmful tech companies are another avenue for action. Writing op-eds, sending letters to the university, and getting news coverage can influence other university constituents, the broader public, and legislators. State-level regulation is ripe for labor intervention, and the AI committee has turned its attention to legislative efforts to protect the rights of all workers in particular states. The connections between education and democracy can be the basis for arguing that laws around educational technology should be made with input from a broader range of constituents, with the involvement of unions and collective action among academic workers and students who are the experts on these technologies. In these areas of activity, higher education labor organizations can work at “bringing the fragments together” in organizing to assert power over the governance of ed-tech and AI as part of the broader fight to protect academic freedom and working conditions. It is crucial that this be a cross-sector, and cross-institutional, organizing process that regularly demonstrates worker power and support for alternative approaches to technology governance. This is how we win.
Britt Paris is associate professor in the School of Communication and Information at Rutgers University. She is on the Rutgers AAUP-AFT Faculty Executive Council and is chair of the national AAUP’s ad hoc Committee on Artificial Intelligence in Academic Professions. Her new book is Radical Infrastructure: Imagining the Internet from the Ground Up. Rebecca Reynolds is associate professor in the School of Communication and Information at Rutgers and a member of the ad hoc Committee on Artificial Intelligence in Academic Professions. Her work investigates intentional and incidental human learning in formal educational settings and informal online settings. She is the cofounder and editor in chief of the academic journal Information and Learning Sciences.