University of Alabama's Guidelines for Responsible AI-Generated Media Use
In-depth discussion
Formal and directive
0 0 3
This document outlines the University of Alabama's guidelines for the ethical and responsible use of AI-generated media. It emphasizes authenticity, credibility, security, and brand protection. Key restrictions include not publishing unedited AI content as finished work, avoiding misrepresentation of real people, places, or events, and never using AI to replicate official brand marks. Transparency through clear labeling is mandated, and inputting confidential information into AI tools is prohibited. The guidelines also stress the need for human review of AI-generated accessibility features and caution against copyright issues.
main points
unique insights
practical applications
key topics
key insights
learning outcomes
• main points
1
Provides clear and actionable guidelines for AI media usage.
2
Emphasizes ethical considerations and brand integrity.
3
Addresses critical aspects like transparency, data security, and copyright.
• unique insights
1
Specific prohibitions against using AI to represent real UA entities (people, places, events).
2
Detailed examples of what constitutes misuse and proper disclosure.
• practical applications
Offers essential rules and best practices for any individual or department within the University of Alabama needing to use AI-generated media, ensuring compliance and maintaining institutional standards.
• key topics
1
AI-generated media ethics
2
University branding and intellectual property
3
Transparency and disclosure requirements
• key insights
1
Establishes specific institutional policies for AI media, tailored to the University of Alabama.
2
Provides clear boundaries on what is permissible and what is not when using generative AI.
3
Highlights the importance of human oversight and ethical conduct in AI content creation.
• learning outcomes
1
Understand the ethical implications of using AI-generated media.
2
Learn to identify and avoid misuse of AI content in official communications.
3
Comply with institutional policies regarding AI media creation and dissemination.
At the heart of the University of Alabama's approach to AI-generated media are three fundamental principles: Authenticity and Credibility, Security, Privacy and Compliance, and Protection of UA Brand and Intellectual Property. As a public research institution, UA prioritizes communications and media that are authentic, truthful, and serve the public interest. This means ensuring that any AI-assisted content accurately reflects reality and does not mislead the audience. Security, privacy, and compliance are paramount, requiring the protection of sensitive institutional and personal information from external systems and strict adherence to all federal, state, and local laws. Finally, the use of AI technologies must enhance, not detract from, the UA brand's storytelling and visual identity, preserving its integrity and consistency.
“ Content Restrictions and Prohibitions
To provide clarity, the guidelines offer specific examples of prohibited AI usage. For instance, if generative AI content is created for illustrative purposes depicting UA traditions like commencement ceremonies or sporting events, it must be clearly labeled as AI-generated, illustrative, or speculative. Publishing AI-generated portraits of UA students, faculty, staff, athletes, or administrators is forbidden. Similarly, using AI-generated images or videos of iconic campus landmarks, such as Denny Chimes or Bryant-Denny Stadium, in official promotional materials is not permitted. Altering visual content to misrepresent its original meaning is also strictly forbidden. This includes adding or removing people from event photographs to imply attendance or absence, altering campus imagery to suggest non-existent facilities, or fabricating visual scenarios that falsely depict interactions between individuals.
“ Brand Mark Protection and AI
Transparency is a cornerstone of the University of Alabama's AI media policy. Any published content that is generated or substantially influenced by AI must include an explicit and clearly visible disclosure to the audience. This disclosure is required regardless of the AI's role, whether it's generating illustrations, summaries, or reports. For example, AI-generated illustrations or visuals should be accompanied by a brief disclaimer, and summaries or reports produced with AI assistance should be noted. Minor AI-assisted edits, similar to traditional digital photo editing tools (e.g., white balance, color rebalance), do not necessitate disclosure. A sample disclaimer for an AI-generated image of an elephant covered in circuitry for a UA cybersecurity news story would be: 'Generative AI tools were used to create this image.'
“ Data Security and Confidentiality with AI Tools
While AI tools can offer valuable assistance in creating accessible content, the University of Alabama emphasizes that human oversight remains indispensable. AI can aid in generating alt-text for images, captions for videos, and transcriptions for audio content. However, users must always review these AI-generated features to ensure their accuracy and appropriateness for all users. For instance, AI-generated captions for videos must be manually verified for accuracy before publication, and AI-generated alt-text for images should be edited to accurately reflect the context and relevance of the visual content. Relying exclusively on AI-generated accessibility features without human review is not permitted.
“ Ethical Considerations and Responsible AI Use
Users of AI-generated content should be aware of the potential implications for copyright protection. The U.S. Copyright Office has indicated that content generated entirely by AI, without meaningful human creative input, may not qualify for copyright protection. Additionally, generative AI tools can inadvertently reproduce elements of copyrighted works from their training data. Therefore, teams should carefully weigh these factors when deciding where to use AI-generated material versus original human-created work, especially for high-value or long-lived assets. This means that AI-generated text, images, or other media used in UA materials may not be legally protectable as UA intellectual property. Caution is advised when using AI-generated content that closely resembles existing published works, designs, or visual styles. For flagship campaigns, signature creative assets, or materials with a long shelf life, prioritizing original human-created work is strongly recommended.
We use cookies that are essential for our site to work. To improve our site, we would like to use additional cookies to help us understand how visitors use it, measure traffic to our site from social media platforms and to personalise your experience. Some of the cookies that we use are provided by third parties. To accept all cookies click ‘Accept’. To reject all optional cookies click ‘Reject’.
Comment(0)