Search Take the survey

↑ ↓ choose a page · Enter searches everything, with exact title matches first · kind:source and other filters work there · Esc closes

← raw/legislative
statute
First page of the original document

2026-11-09 AB 2392 Assembly committee analysis

Original ↗ Online ↗

url
https://billtexts.s3.amazonaws.com/ca/ca-analysishttps-leginfo-legislature-ca-gov-faces-billAnalysisClient-xhtml-bill-id-202520260AB2392-ca-analysis-399036.pdf
fetched
2026-11-09
publisher
California State Assembly
published
unknown
original
2026-11-09 AB 2392 Assembly committee analysis.pdf
rehost
yes

Links to 0 · Referenced by 1

Markdown copy of 2026-11-09 AB 2392 Assembly committee analysis.pdf, made 2026-11-09. The original is the authority. AB 2392 Page 1

ASSEMBLY THIRD READING AB 2392 (Fong) As Amended April 23, 2026 Majority vote

SUMMARY

Requires the California Community Colleges (CCC) and the California State University (CSU), and requests the University of California (UC), to convene a joint working group to develop responsible generative artificial intelligence (GenAI) training and procurement standards and present its recommendations, by January 1, 2028, to the respective public higher education system leaders. Requires the CCC and CSU, and requests the UC, to provide a training developed by the working group to students, faculty, and staff prior to their use of any procured GenAI system, to review that training at least once per academic year, and to maintain records of completed trainings for a minimum of four years.

Major Provisions

  1. Requires the CCC and CSU, and requests the UC, before providing a GenAI system to students, faculty, or staff, to jointly convene a working group to develop responsible standards for GenAI training and procurement, including:

a) Procurement standards ensuring that any procured GenAI system meets specified

minimum protections relating to risk assessment, sycophancy, safety monitoring, incident reporting, data governance, data minimization, training data disclosure, vendor contracting, and labor practices; and,

b) A training on the use of any procured GenAI system covering institutional policies,

including academic integrity policies, applicable privacy disclosures, and guidance on the limitations of GenAI systems, including the potential for inaccurate, incomplete, or misleading outputs. Requires the working group to present its recommendations to the respective system leaders on or before January 1, 2028.

  1. Requires the CCC and CSU, and requests the UC, to provide the training developed by the working group to students, faculty, and staff prior to their use of a GenAI system, and authorizes institutions to develop distinct versions of the training tailored to the distinct uses, responsibilities, and applicable policies relevant to students, faculty, and staff, respectively.

  2. Requires the CCC and CSU, and requests the UC, to review the training at least once per

academic year and update it as necessary to reflect changes in institutional policy, applicable law, or the functionality or privacy practices of the GenAI system, and to provide updated trainings to users before their continued use of a product that has undergone a material change in functionality or privacy practices.

  1. Requires the CCC and CSU, and requests the UC, to maintain records of training completion for each student, faculty member, and staff member who receives a training, including the name of the individual, the date of completion, and the version of training delivered, and to retain those records for a minimum of four years following the date of completion.

  2. Requires the CCC and CSU, and requests the UC, to submit a written report to the

Legislature and relevant policy committees with jurisdiction over higher education and

AB 2392 Page 2

privacy and consumer protection, as specified, before executing any contract for the procurement of a GenAI system during the period beginning January 1, 2027, and ending upon the adoption of procurement standards by the working group. Requires the report to include a description of the GenAI system to be procured, the process used in evaluating and selecting the system, the training to be provided prior to deployment, and the vendor’s data use, retention, and privacy practices.

  1. Makes the following definitions:

a) Defines “generative artificial intelligence” or “GenAI” and “artificial intelligence” by

reference the Government Code, as specified, and,

b) Defines “sycophancy” as the tendency of a GenAI system to overly agree with, validate, or flatter a user even when the user is communicating factually inaccurate or illegal ideas or actions for the primary purpose of optimizing engagement.

COMMENTS

CSU AI-Empowered Initiative. On February 4th, 2025, the CSU system announced their AI- Empowered Initiative. In their press release, the CSU indicated that their goal was to “leverage the power of artificial intelligence to create an AI-empowered higher education system that could surpass any existing model in both scale and impact.” Among the various elements of this initiative included the systemwide deployment of an AI tool developed by OpenAI - ChatGPT Edu – that would enable “…students, faculty, and staff to use AI to accelerate learning, optimize workflow efficiency, and foster cross-departmental collaboration.” The contract for ChatGPT Edu $16.9 million, and runs from February 2025 through July 2026.

Although “ChatGPT” has become a somewhat ubiquitous term for generative AI tools, many different technology companies have deployed AI products that are now embedded in common e-mail, word processing, and document production systems. An element of CSU’s AI- Empowered Initiative also highlights partnerships with some world’s leading tech companies, including Adobe, Alphabet (Google), AWS, IBM, Instructure, Intel, LinkedIn, Microsoft, and NVIDIA, in addition to OpenAI.

Student response. The California State Student Association (CSSA) released a white paper regarding the CSU’s AI-Empowered Initiative on February 24th, 2026. The whitepaper noted that, while students broadly recognize AI’s potential to enhance creativity and learning, the implementation has been rushed, inconsistent, and lacking adequate training and awareness. Many students are unaware of the tool or unsure how to use it, and there is no unified guidance across classrooms, leading to confusion about acceptable AI use. This inconsistency is compounded by conflicting faculty approaches—some encourage AI while others penalize it— along with unreliable AI-detection tools that have led to false accusations and increased anxiety.

Beyond implementation challenges, the white paper highlights major concerns about privacy, transparency, sustainability, and governance. Students worry about how their data is collected and used, expressing distrust in the system despite the availability of an official platform. Students also question the financial transparency of the initiative and its alignment with broader CSU budget cuts. Environmental concerns are prominent as well, with many students criticizing the lack of attention to AI’s energy and water consumption and its conflict with CSU’s climate goals.

AB 2392 Page 3

Additionally, the paper raises ethical and mental health concerns, including risks from harmful AI outputs and overreliance on the technology. Across all areas, a central theme emerges: students feel excluded from decision-making and call for greater oversight, clearer policies, and consistent inclusion in governance to ensure AI is implemented responsibly, ethically, and in alignment with student needs.

Joint Hearing. In order to better understand the specifics of CSU’s AI-Empowered Initiative, the Assembly Committee on Higher Education partnered with the Assembly Committee on Privacy and Consumer Protection to conduct an Oversight Hearing on August 26, 2025.

Upon questioning from the Committees, it became clear that there had been no training requirement in place for students, faculty, and staff prior to the deployment of ChatGPT Edu. Testimony also indicated confusion from students, faculty, and staff regarding specific institutional AI policies, from general privacy concerns to the degree with which generative AI products can be used in academic and workplace settings.

Following the hearing, the CSU Chancellor’s Office sent follow-up information to the Committees to address specific about required training. CSU noted that the system “…is committed to providing its students and employees with training opportunities to best use the technology platforms it makes available. While training on AI or other technology platforms is not required, many campuses have begun including the use of AI into their new student orientation and first-year experience activities. Additionally, a variety of online and in-person AI training courses are offered to students, staff and faculty via the Chancellor’s Office and individual campuses. The CSU hosts the AI Commons training website that has a wide range of training opportunities.”

Existing workgroups and initiatives. The UC, CSU, and CCCs have each undertaken AI-related initiatives. Examples include:

  1. The UC convened the “UC Presidential Working Group on Artificial Intelligence” in 2020,

which issued a final report in 2021 titled “Responsible Artificial Intelligence: Recommendations to Guide the University of California’s Artificial Intelligence Strategy.” The recommendations included in this final report include:

a) Institutionalize the UC Responsible AI Principles in procurement, development,

implementation, and monitoring practices;

b) Establish campus-level councils and support coordination across UC that will further the

principles and guidance developed by this Working Group;

c) Develop an AI risk and impact assessment strategy; and,

d) Document AI-enabled technologies in a public database.

  1. As part of their AI-Empowered Initiative, the CSU announced in February 2025 that it would be using a comprehensive strategy, which included an AI Workforce Acceleration Board. The CSU Board of Trustees Joint Committee on Educational Policy & Finance presented the CSU Artificial Intelligence Strategy during the Board’s January 2025 meeting. The Academic Senate of the CSU called for a working group on AI in higher education in 2023, and has passed subsequent resolutions regarding AI in 2024 and 2025.

AB 2392 Page 4

  1. The CCC has established the Digital Center for Innovation, Transformation and Equity to represent a strategic approach to supporting the CCC’s Vision 2030, including leading in innovation, harnessing the power of generative AI, public and private technology partnerships, and transforming the CCC’s ability to serve students as they move into an AI- powered workforce.

Additionally, in November 2024, the California Education Learning Lab announced the recipients of the “AI Grand Challenge: Leveraging AI for Teaching and Learning” grants, aimed at fostering innovation in the use of AI to enhance curriculum and pedagogy within California’s public higher education system. Nearly 75 institutions from the CCC, CSU and UC systems submitted 32 proposals. Five projects have been selected to receive grants of up to $1.5 million each. The Foundation for California Community Colleges is administering the grants on behalf of the State. The awarded projects aim to create significant influence on teaching and learning experiences statewide and will strive to reach approximately 150,000 students and 13,500 faculty/instructors across California’s public higher education institutions and beyond.

According to the Author According to the author, “On August 26th, 2025, the Assembly Committees on Higher Education & Privacy and Consumer Protection jointly held an oversight hearing to examine the California State University’s AI-Empowered Initiative. A key finding of this hearing was that students, faculty, and staff did not receive training on ChatGPT Edu prior to the deployment of that tool throughout the CSU system. Another key finding of the hearing was that students, faculty, and staff expressed a lack of clarity around institutional policies concerning AI use, and the privacy of information shared with AI systems.”

The author continues that, “AB 2392 applies lessons learned from the rollout of CSU’s AI- Empowered Initiative, and will ensure that public segments of higher education, before providing a GenAI system to students, faculty, or staff provide training on the use of that product. AB 2392 also ensures that training includes relevant policies concerning the use of artificial intelligence, and a disclosure of privacy policies for the GenAI product.”

Arguments in Support The California Teachers Association (CTA) wrote in support of the measure, noting that, “AI tools, programs, and applications are becoming increasingly pervasive in California’s public higher education systems. This underscores the critical need for educators, students, and staff to be adequately trained on their uses and implications. AB 2392 would establish a necessary framework for transparency and labor protection in AI’s application.”

CTA continues that, “as campuses explore AI platforms to track student engagement (e.g., logins, discussion frequency, or assignment completion), there is an increasing risk that these data points will be used to generate instructor effectiveness reports. Without the training mandated by AB 2392, faculty may be evaluated based on opaque, algorithmic metrics they do not understand and cannot challenge. The bill ensures that training covers the use of the product, which must include a clear explanation of how metrics are calculated and weighted.”

CTA adds that, “additionally, in many scenarios, institutions may adopt AI tutoring systems without a clear understanding of how student learning data is handled. This risks sensitive student data being sold to or utilized by third-party vendors without meaningful disclosure. AB 2392 mandates that training include relevant policies and privacy disclosures, which ensures that

the campus community understands exactly how their data is being handled. This helps prevent privacy breaches and empowers faculty to guide students on the limitations of these tools.”

AB 2392 Page 5

Arguments in Opposition None on file.

FISCAL COMMENTS

One-time General Fund costs of an unknown amount, likely minor and absorbable, to the CCC, CSU, and UC to convene the working group and provide recommendations.

VOTES

ASM HIGHER EDUCATION: 10-0-0 YES: Fong, DeMaio, Boerner, Jeff Gonzalez, Jackson, Irwin, Patel, Bennett, Sharp-Collins, Tangipa

ASM PRIVACY AND CONSUMER PROTECTION: 15-0-0 YES: Bauer-Kahan, Macedo, Bryan, DeMaio, Hoover, Irwin, Lowenthal, McKinnor, Ortega, Patterson, Pellerin, Petrie-Norris, Ward, Wicks, Wilson

ASM APPROPRIATIONS: 14-0-1 YES: Wicks, Hoover, Aguiar-Curry, Calderon, Caloza, Dixon, Fong, Mark González, Krell, Pacheco, Pellerin, Sharp-Collins, Solache, Ta ABS, ABST OR NV: Tangipa

UPDATED

VERSION: April 23, 2026

CONSULTANT: Kevin J. Powers / HIGHER ED. / (916) 319-3960

FN: 0002952

Suggest a change to this page

Source file: raw/legislative/2026-11-09 AB 2392 Assembly committee analysis.md.

Suggest or correct

Know a report, dataset or law that belongs here? Found something wrong? Have a question the sources could answer? A consortium member reads every suggestion and decides. Nothing changes until a person makes the change, and the Suggestions page shows what happened to it.

About you (optional)

Your name and email stay with the team; we use the email only to reply. Your institution may appear on the contributors board.