AI-AI… Oh! Considerations for Using AI-Generated Infographics

I was wondering whether the use of infographics created using Open AI would be acceptable in the presentation?

​One of my graduate educational technology students just emailed to ask me this question. I’m glad they did, because it shows that they are thinking about what digital tools they are going to use, and how they are going to use them as part of their coursework. Normally, I’d send off a quick reply. But, the question of the acceptability of using an AI-generated infographic does not have a simple yes/no answer. 

​The truth is, there are a number of things to consider related to several key ethical issues, three of which are:

• Academic Integrity
• Accessibility
• Sustainability

---

Academic Integrity

​The first question to ask is whether you are even permitted to use AI tools to generate content for your assignment. Does your instructor permit them? What about your school’s academic integrity or AI usage policies? Some schools ban the use of AI outright. Others, such as Cape Breton University (2026a, b) and Ontario Tech University (2026) have published frameworks or policies about what represents acceptable or responsible use. In some cases, those schools’ Centers for Teaching and Learning offer both instructors and students useful guidelines on how to ethically incorporate AI tools into assignments or projects. 

For instance, CBU’s CTL (n.d.) promotes a Stoplight (Red Light/Green Light/Amber Light) decision-making matrix, highlighting whether the use of AI tools is off-limits, fair game, or if students should proceed with caution. Ontario Tech University’s Teaching and Learning Center (2026) provides a detailed numerical level framework, ranging from Level 1 (no AI use) to Level 6 (Full AI – AI is integral to the creation of work). In the cases of both CBU and Ontario Tech University, instructors are encouraged to include clear statements on their syllabi as to what constitutes acceptable AI use. Ontario Tech University also now encourages instructors to indicate the AI in Assessment Scale level designated for each individual assignment both in their syllabi and detailed assignment instructions. ​

Figure 1: OnTechU's AI in Assessment Scale (Ontario Tech Univeristy, 2026)

For simplicity’s sake, let’s assume that your instructor has designated your assignment as either “Green Light” or “Level 6 – Full AI.” You are permitted to leverage AI tools to generate materials for your assignment or project. That doesn’t mean you can just proceed without further due diligence to academic integrity! My general rule when using AI tools is that I should be able to stand behind the content. I should have confidence in its veracity, and I should be able to say that I used the AI tools to produce resources representing my work. I would never use AI tools to generate the content for me. 

• I did the knowledge curation or generation.
• AI produced resources based on my content.
• I did not use AI to generate content.
• I have verified the veracity of any resources cited within the resource, as well as the messages conveyed about those resources.

Biases

Returning to my student’s original inquiry, I would personally avoid the use of OpenAI to create presentation resources or infographics. Platforms based on OpenAI such as ChatGPT (OpenAI, n.d.), like many Large Language Model (LLM) tools, have inherent biases in the content that they generate (Power, 2024d, 2025). Without careful prompting and a strong personal grounding in the subject matter you are working with, resources that you generate with AI tools may inadvertently promote dominant cultural narratives and biases at the expense of marginalized perspectives. It is your responsibility to question these narratives and to present accurate, unbiased information. 

On strategy that I use to mitigate the biases that may result from trawling the open web for content, using skewed algorithms, is to restrict my resource generation activities to Google’s (2026b) NotebookLM. NotebookLM will only generate notes, slide decks, images, podcasts, or videos, based on content that I have curated. While the LLM backbone of NotebookLM may still have biases within its algorithms, at least I can be confident in the veracity of the source materials that I provide to it. BUT, I must still be careful with my prompting. AND, I must still carefully review anything I have asked it to generate for me. I need to be able to stand behind the messages conveyed by those resources, as much as I need to be able to stand behind the sources of those messages! If I’m not confident that the messaging is accurate, unbiased, and free from the propagation of stereotypes, then I need to go “back to the drawing board.” I either need to manually edit the content, or revise my prompting strategies to get the tools to generate something more acceptable.

Cite It!

If you do choose to use AI tools to help you produce content for assignments, projects, or presentations, be sure to cite the tools you’ve used and the content they’ve generated appropriately. CBU’s Academic Integrity Handbook (2020) provides some excellent resources to help you with properly citing AI-generated content.

Intellectual Property Rights

Issues of academic integrity go beyond honesty in one’s own work. They also encompass the promotion and protection of intellectual property rights. Faculty and students must both critically evaluate the impact of Generative AI applications on intellectual property rights on a global scale. There are serious concerns around how AI tools have been “trained” through the unauthorized use (and in many cases remixing and redistribution) of copyright-protected materials (as well as personal information) (Di Placido, 2025; Gillham, 2024; Heydari, 2024; Steinberger, 2024). Before leveraging AI tools for any otherwise well-intentioned, field-specific purpose, faculty and students should question whether the benefits of using such tools outweigh the harms to their colleagues and others whose work may have been misappropriated.

---

Accessibility and Infographics

Confidence in the academic integrity of resources that you have used AI tools to produce for you is not enough. Using infographics to convey information is pointless of your audience cannot access that information!

Text-Within-An-Image

Too often, I come across infographics that rely too heavily on text. Remember, an infographic is essentially a digital poster. An image. And text-within-an-image is inherently inaccessible. It is not machine-readable. Therefore, it cannot be accessed using a screen reader application. If your content is primarily text, then an infographic is not the best choice for presenting it. 

If you are using the infographic as part of a live presentation, including too much text creates two significant risks (Hayes, 2025; Presono, 2022; Rivera, n.d.). One risk is that you will end up simply reading that text word-for-word – which makes for a poor and unengaging presentation. If you know the content well enough that you don’t read the on-screen text word-for-word, you create another accessibility and engagement issue – your audience will experience cognitive overload trying to follow both your speaking and the on-screen text! 

If you are using your infographic as media content in an asynchronous presentation (such as embedding it on a group presentation website), then the inaccessibility of text-within-an-image becomes a real problem for you and your audience (Government of Canada, 2025; University of South Carolina, n.d.). You must ensure that the embedded image contains appropriate ALT text that describes the content and purpose of that image (without repeating the text word-for-word). And, because the text content is the primary focus of the image, you need to repeat those key points in actual machine-readable text as part of the web page content. If that sounds kind of redundant and pointless, it is! A good infographic should focus on image-based information, with text minimized to content headings and a few key points. The infographic should reinforce the content you are presenting on your web page. It should NOT replace it! 

For more on these issues, refer to A Picture Isn’t Always Worth a Thousand Words… (Power, 2024a).

Alternate Formats

Let’s say that you DO want to use your infographic as the primary means of conveying content information to the audience of your web page. Well, since your infographic is an image, you will need to provide an alternate, accessible version of that poster (Government of Canada, 2025; University of South Carolina, n.d.). The easiest way to do that is to provide a link to a PDF version of the infographic. But, that raises a whole range of additional considerations to ensure that the PDF, itself, meets document accessibility standards, including (but not limited to):

• Proper tagging of content headings and paragraph text.
• Inclusion of appropriate ALT text for graphics or images within the infographic that have instructional value.
• Ensuring that embedded graphics or images that are purely decorative are actually tagged as such (so that a screen reader will ignore them).
• Verifying the reading order of all objects (text and graphics) on your infographic, so that they are read out in the intended sequence to anyone using a screen reader.

For more on this, refer to Are Your PDFs Accessible? (Power, 2024c) and Accessible Presentations (Power, 2024b).

Color

So, you’ve decided to continue with the use of an infographic and you’ve sorted out the balance of text and graphics, and alternate means of conveying the key content to your audience. That still does not guarantee that your infographic is accessible! Color choices for your poster are more than aesthetic decisions. Yes, you should pay attention to color choice to maximize engagement and impact (Bandli, 2025; Malamed, n.d.). But, you also need to ensure your color choices do not create barriers for your audience. 

• Is text clearly distinguishable from the background (color contrast ratio)?
• What happens to your content if it is viewed in dark mode or grey-scale? Or if it is printed in grey-scale and photocopied?
• Have you relied only on color to convey key information or concepts? 

For more on this, refer to Should You Use Color for Emphasis? (Power, 2024e)

---

Sustainability

Now that you’ve addressed academic integrity concerns and the ability of your audience to access your infographic, there are some additional ethical considerations that you should be aware of. Long story short, you need to ask yourself whether the benefits of using AI tools to generate that infographic are worth the costs.

Besides the immediate costs of subscriptions for faculty, students, and institutions, there are more global costs that bring into question the sustainability of widespread use of AI applications. Steinberger (2025), Tadorovic (2024), and Wright (2025) all emphasize the disproportionate energy use, water use, and carbon footprint of Generative AI. Wright (2025) notes that using ChatGPT just to draft a 100-word email "sweats off more than an Evian bottle’s worth of water (519 millilitres)... [and] the electricity used to generate that 100-word email is equal to powering 14 LED light bulbs for an hour (0.14 kilowatt-hours (kWh))." According to Goldman Sachs (2024), “a ChatGPT query needs nearly 10 times as much electricity to process as a Google search,” and the growing demand for AI use may lead to “data center power demand... grow[th of] 160% by 2030.” Such high natural resource demands may be unsustainable, and antithetical to strategic goals of your school or organization. In light of this, when considering how to integrate AI applications, and demonstrate their ethical use, you should carefully consider whether the benefits of using such tools are sustainable and outweigh their potential environmental impacts.

---

An Infographic Example

An AI-Generated Infographic About Using AI-Generated Infographics

​This summary infographic was created using NotebookLM with the following prompt:

​Generate an infographic summarizing the key considerations for using an AI-generated infographic in an assignment or project, based on the document AI AI Oh. Use a blue and white color scheme, with a sufficient color-contrast ratio to meet WCAG 2.2 AA criteria for readability. Use minimal text as headings and to summarize key points. Be sure to separate the categories of Academic Integrity, Accessibility, and Sustainability.​

Figure 2: Ethical Infographics in AI Age (Google, 2026a)

Analysis of the AI-Generated Infographic

​In the spirit of "standing behind" my own work, and resoruces that I use AI tools to generate, here is my assessment of this AI-generated infographic:

• It meets my academic integrity requirements. (It was generated by NotebookLM exclusively from the text of this blog post, which I researched and authored.)
• It meets WCAG 2.2 color-contrast AA requirements for text readability (World Wide Web Consortium, 2024).
• It contains appropriate ALT text for the infographic as a whole (which I manually added when uploading the image for this post).
• It still contains a fairly large amount of text-within-an-image, which fails to meet image accessibility goals.
• The text could not be detected using optical character recognition (OCR), so it was not possible to produce an accessible PDF version of the infographic for sharing in this post. (However, ALL of the key points within the infographic are explained in detail in the main body of this post.)
• There is no ALT text for the individual graphics embedded in the infographic (as there is no accessible PDF version of the infographic.

The Verdict

Despite the identified accessibility issues, I would feel confident standing by the use of this infographic in terms of both academic integrity and general accessibility. This infographic would be suitable for use in a live presentation, as I would not be reading the on-screen text word-for-word. It is also acceptable to use in an asynchronous presentation such as this blog post, as the key points are available in an alternate format in the main text of the post, and ALT text is available for the infographic image. I have also provided citations in this post for both the AI tool used to generate the image, and the AI-generated image itself.

The only considerations that I'm not truly satisfied with are the economic and environmental issues. While it might take me more time, I think that I would still use non-AI tools to create these types of infographics for my presentations.

---

References

Bandli, S. (2025, August 27). Understanding Colour Theory for Learning Experience Design. [Web log post]. Who's Your ADDIE? https://www.whosyouraddie.com/post/colour-theory-for-learning-experience-design

​Cape Breton University (2020). 6.3 Citing AI. Academic Integrity Handbook. https://pressbooks.atlanticoer-relatlantique.ca/acadintegritycbu/chapter/citing-ai/

Cape Breton University (2026a). Academic Integrity. https://www.cbu.ca/current-students/academic-integrity/

Cape Breton University (2026b). Artificial Intelligence at Cape Breton University. https://www.cbu.ca/about-cbu/artificial-intelligence-at-cape-breton-university/

CBU Library (n.d.). How to use and NOT use AI Tools. https://libguides.cbu.ca/c.php?g=734809&p=5287845

Di Placido, D. (2025, March 27). The AI-Generated Studio Ghibli Trend, Explained. Forbes. https://www.forbes.com/sites/danidiplacido/2025/03/27/the-ai-generated-studio-ghibli-trend-explained/

Gillham, J. (2024, December 17). OpenAI and ChatGPT Lawsuit List. Originality.ai. https://originality.ai/blog/openai-chatgpt-lawsuit-list

Goldman Sachs (2024, May 14). AI is poised to drive 160% increase in data center power demand. https://www.goldmansachs.com/insights/articles/AI-poised-to-drive-160-increase-in-power-demand

Google (2026a, March 17). Ethical Infographics in AI Age. [AI-generated infographic]. NotebookLM [Image generator]. https://notebooklm.google.com/

Google (2026b). NotebookLM [Small Language Model AI application]. https://notebooklm.google.com/

Government of Canada (2025, September 19). Alternative text and long description - Best practices. https://a11y.canada.ca/en/alternative-text-and-long-description-best-practices/

Hayes, M. (2025, January 13). What the Research Says About How Text-Heavy Slides Impact Investor Day Audiences. [Web log post]. LinkedIn Pulse. https://www.linkedin.com/pulse/what-research-says-how-text-heavy-slides-impact-investor-mark-hayes-756ue

Heydari, A. (2024, November 9). Media outlets, including CBC, sue ChatGPT creator. CBC News. https://www.cbc.ca/news/business/openai-canadian-lawsuit-1.7396940

Malamed, C. (n.d). What Learning Professionals Should Know About Color. [Web log post]. The eLearning Coach. https://theelearningcoach.com/media/graphics/graphics-primer-color/

Ontario Tech University (2026). AI in Assessment Scale. https://tlc.ontariotechu.ca/nexthub/artificial-intelligence/ai-scale/

OpenAI (n.d.). ChatGPT. [Software application]. https://chat.openai.com

Power, R. (2024a). A Picture Isn’t Always Worth a Thousand Words… In The ALT Text: Accessible Learning with Technology. Power Learning Solutions. https://pressbooks.pub/thealttext/chapter/a-picture-isnt-always-worth-a-thousand-words/

Power, R. (2024b). Accessible Presentations. In The ALT Text: Accessible Learning with Technology. Power Learning Solutions. https://pressbooks.pub/thealttext/chapter/accessible-presentations/

Power, R. (2024c). Are your PDFs Accessible? In The ALT Text: Accessible Learning with Technology. Power Learning Solutions. https://pressbooks.pub/thealttext/chapter/are-your-pdfs-accessible/

Power, R. (2024d). Evaluating Graduate Education Students’ Self-Efficacy with the Use of Artificial Intelligence Agents. Journal of Educational Informatics, 5(1), 3-19. https://journalofeducationalinformatics.ca/index.php/JEI/article/view/269 

Power, R. (2024e). Should You Use Color for Emphasis? In The ALT Text: Accessible Learning with Technology. Power Learning Solutions. https://pressbooks.pub/thealttext/chapter/should-you-use-color-for-emphasis/

Power, R. (2025). Seamlessly Integrating and Evaluating AI Competencies for Graduate Education Students. In Hambrock, H., & Power, R. (Eds), Artificial Intelligence for Seamless Education. International Research Network for Innovative Sustainable Seamless Education and Power Learning Solutions. ISBN 978-1-7390190-3-7. https://pressbooks.pub/aiforseamlesseducation/chapter/thechattproject/ 

Presono. (2022, May 13). This is the real reason why too much text on slides doesn't work. https://presono.com/en/blog/2022-05-13-11-35-06/

Rivera, E. (n.d.). You. Yes, you. You're using too much text on your presentation slides! [Web log post]. Dr. Echo Rivera. https://www.echorivera.com/blog/toomuchtext

Steinberger, J. (2025). “Sharing the slides I used to try to convince my students NOT to use Gen AI, aka artificial stupidity.” [LinkedIn post]. https://www.linkedin.com/posts/julia-steinberger-69139b1_sharing-the-slides-i-used-to-try-to-convince-activity-7305540616610852865-WRoS?utm_source=share&utm_medium=member_desktop&rcm=ACoAAALoqOwBDiJDl830ePQ_JLQsTe9bOxP94Bk

Tadorovic, I. (2024, September 4). ChatGPT consumes enough power in one year to charge over three million electric cars. Balkan Green Energy News. https://balkangreenenergynews.com/chatgpt-consumes-enough-power-in-one-year-to-charge-over-three-million-electric-cars/

World Wide Web Consortium (2024, December 12). Web Content Accessibility Guidelines (WCAG) 2.2. https://www.w3.org/TR/WCAG22/

Wright, I. (2025, February 17). ChatGPT Energy Consumption Visualized. Business Energy UK. https://www.businessenergyuk.com/knowledge-hub/chatgpt-energy-consumption-visualized/

​University of South Carolina (n.d.). Infographic ALT Text. Digital Accessibility. https://sc.edu/about/offices_and_divisions/digital-accessibility/toolbox/best_practices/alternative_text/infographic-alt-text