Prohibited AI Models and Technologies

Summary

At Florida Atlantic University, data privacy, intellectual property protection, and regulatory compliance are critical to our research and educational missions. You may have noticed that certain artificial intelligence (AI) models and applications are completely blocked on the campus network or prohibited on university-issued devices.

To keep it brief: FAU prohibits the use of AI models and software applications developed or owned by entities located in designated "Foreign Countries of Concer

Body

Overview

At Florida Atlantic University, data privacy, intellectual property protection, and regulatory compliance are critical to our research and educational missions. You may have noticed that certain artificial intelligence (AI) models and applications are completely blocked on the campus network or prohibited on university-issued devices.

To keep it brief:  The use of AI models developed or hosted or owned by entities located in designated "Foreign Countries of Concern" or on select lists is not allowed.

While some of these models have generated significant global tech headlines, state laws and data security rules are non-negotiable. This article explains the "why" behind the restrictions and lists the specific models you must avoid.

Why are Some AI Models Prohibited at FAU?

The restriction is not an arbitrary IT decision; it is driven by a combination of Florida state legislation, state university system mandates, and severe data privacy concerns:

  • Florida Statute § 287.138 & BOG Regulation 9.012: Florida law strictly prohibits public universities and state agencies from contracting with, utilizing platforms from, or transferring state-funded data to entities controlled by foreign countries of concern (which include China, Russia, Iran, North Korea, Cuba, Venezuela, and Syria).

  • The "National Security Data Leak" Risk: Under Chinese cybersecurity laws, domestic tech companies are legally obligated to share user data, system prompts, and uploaded proprietary code or research with the government upon request. Using these models for university work creates an unacceptable risk of intellectual property theft.

  • State-Level Directives: Florida's Chief Financial Officer and individual State University System (SUS) institutions have explicitly blacklisted specific high-profile Chinese AI platforms from all state-issued devices and networks due to active surveillance and security threats.

Examples of Prohibited AI Models

If an AI model is developed by a company headquartered or heavily controlled within China, it cannot be used for any university-related business. Including Research and Academics. Below is a list of major example models that strictly prohibited. 

This list is incomplete and the absence of a model from this list does not suggest it is approved for use. Under Florida law, any artificial intelligence model or software system owned, developed, or maintained by a foreign principal domiciled within a foreign country of concern is strictly prohibited on state and university networks.

Developer / Company Prohibited AI Model(s) Notes / Context
DeepSeek (High-Flyer Quantum) DeepSeek-V3, DeepSeek-R1 Explicitly blacklisted by Florida state agencies and state university IT networks due to severe data harvesting concerns.
Baidu Ernie Bot, Ernie 4.0, Ernie Speed Baidu’s flagship large language model ecosystem.
Alibaba Cloud Qwen Series (Qwen2.5, Qwen-Audio, Qwen-VL) Includes all open-source and proprietary text, audio, and vision models under the Qwen umbrella.
ByteDance Doubao, Skylark (Doubao-pro) AI systems developed by the parent company of TikTok (which is also banned on FAU devices).
Tencent Tencent Hunyuan (Hunyuan-Large) Enterprise-grade foundational models used for chat and automation.
Zhipu AI GLM, ChatGLM, CogVideoX Widely distributed open-source and commercial foundational models.
Moonshot AI Kimi (Kimi K2.6, Kimi K2.5, Kimi-VL) Advanced long-context and multi-agentic foundational models built by a prominent Beijing-based AI startup.
MiniMax Group Inc. MiniMax, Hailuo (M3.0, M2.7, Hailuo 2.3) Multimodal and text-to-video engines backed heavily by Tencent and Alibaba.
01.AI Yi Series (Yi-Large, Yi-34B) Highly popular open-weights model family restricted due to corporate domicile in Beijing.
SenseTime SenseChat Models built by SenseTime, an organization currently under U.S. trade restrictions.

Source: https://dms-media.ccplatform.net/content/download/163205/file/Prohibited%20Applications%20List_FINAL_4_30_2026.pdf

Compliance Guidelines for the FAU Community

To ensure you stay compliant with state law and university policy, please adhere to the following rules:

⚠️ Important Reminder: Prohibited technologies must not be downloaded, accessed, or executed on any university-owned device (servers, laptops, desktops, tablets, or smartphones).

  1. Do Not Input University Data: Never upload proprietary research, student records, grading rubrics, or unpublished manuscripts into unvetted foreign platforms.

  2. Beware of "Open-Source" Loopholes: Just because a model like Qwen or DeepSeek allows you to download locally does not exempt it from state regulations regarding technology originating from countries of concern.

Safe and Approved Alternatives

You don't have to navigate the AI landscape empty-handed. FAU provides secure, fully vetted alternatives that protect your data:

  • Microsoft Copilot (Enterprise Version): Log in using your FAU single sign-on (@fau.edu) credentials. This enterprise tier features commercial data protection, meaning your prompts, uploads, and responses are completely encrypted and are never used to train external models.

  • Vetted U.S. Implementations: Standard institutional access to tools provided by domestic entities (such as commercial tiers of OpenAI’s ChatGPT, Google’s Gemini, or Anthropic’s Claude) may be used only if they have been reviewed and approved for data compliance by the FAU Office of Information Technology (OIT).

References

Details

Details

Article ID: 162817
Created
Mon 6/8/26 8:52 AM
Modified
Thu 10/8/26 8:31 PM