Transitioning Into the Age of AI: Will Artificial Intelligence Adapt to Human Diversity?

Accessibility, Neurodiversity, and the Hidden Demands of Prompt Literacy

Artificial intelligence is rapidly transitioning from a specialized technological tool into something embedded in everyday life. Students use it to brainstorm essays. Researchers use it to summarize literature. Professionals use it to organize workflows, draft emails, and generate ideas. Increasingly, AI is shaping how we communicate, learn, work, and problem-solve.

Like all major societal transitions, however, the rise of AI raises important questions about accessibility, adaptation, and equity.

As someone interested in stress, mental health, and life transitions, I find myself wondering: How will neurodivergent individuals fare in a world increasingly shaped by AI?

Will AI help level the playing field by reducing barriers related to executive functioning, communication, and information processing?

Or will it create new forms of inequality by rewarding those who can most effectively “speak the language” of AI?

The answer may be more complicated than extreme.

AI as a Potential Equalizer

There is genuine reason for optimism regarding the role AI could play in supporting neurodivergent individuals.

Recent work examining AI and inclusive education suggests these technologies may help personalize learning, improve accessibility, and reduce barriers associated with traditional educational and workplace structures (Dumitru et al., 2026; Melo-López et al., 2025).

AI tools may assist users by:

• breaking complex tasks into manageable steps

• summarizing lengthy information

• organizing ideas into clearer structure

• reducing working memory demands

• assisting with written communication

• offering repetition and clarification without judgment

• helping initiate difficult or overwhelming tasks

For some individuals with ADHD, AI may function almost like an “external executive functioning support,” helping organize thoughts and reduce cognitive overload (Ronksley-Pavia et al., 2025). For some autistic individuals, AI may provide a less socially demanding space to clarify ideas, rehearse communication, or navigate ambiguous expectations (Ronksley-Pavia et al., 2025).

In this sense, AI has the potential to become an important equalizing technology.

Historically, many neurodivergent individuals have experienced a disconnect between what they know and what they are able to demonstrate (Tenorio et al., 2014) within systems that privilege speed, organization, verbal fluency, or conventional communication styles (Ogut et al., 2025). AI may help bridge some of these gaps (Holmes et al., 2019).

At the same time, neurodivergent individuals may also bring strengths particularly valuable in AI interaction, including:

• systems thinking

• persistence

• pattern recognition

• attention to detail

• unconventional associations

• analytical problem-solving

Some individuals may become exceptionally sophisticated AI users precisely because they approach technology differently (Riefle et al., 2022).

But AI Also Introduces New Cognitive Demands

Transitions rarely eliminate barriers entirely. More often, they replace old barriers with new ones.

One emerging concept is what some are beginning to call “prompt literacy.”

Prompt literacy refers to the ability to effectively communicate with AI systems by:

• asking clear questions

• refining requests

• providing sufficient context

• recognizing when outputs are incomplete or inaccurate

• reworking prompts iteratively

• translating goals into language the system can interpret effectively

In many ways, prompt literacy may become a new form of digital literacy (Hwang et al., 2023).

The challenge is that effective prompting often depends upon skills that are not evenly distributed across individuals, including:

• abstract reasoning

• verbal precision

• self-awareness of one’s goals

• cognitive flexibility

• tolerance for ambiguity and trial-and-error

• task decomposition

For some people, this process feels intuitive. For others, it can become cognitively exhausting very quickly.

When AI Still Requires Too Much Cognitive Labour

This tension became personally meaningful for me while observing my neurodivergent adult stepdaughter attempt to use ChatGPT.

Initially, I wondered whether AI might become especially helpful for her, particularly given some of the executive functioning and communication challenges she experiences. In theory, AI seemed like a tool that could help organize information, answer questions, and reduce frustration.

In practice, however, we encountered an unexpected barrier.

Often, she knows she wants help with something but has difficulty translating that internal need into sufficiently precise language. Effective AI use frequently requires users to:

• clarify what they are seeking

• identify missing information

• refine prompts repeatedly

• reinterpret vague responses

• communicate intentions with increasing specificity

For many users, these steps become part of an iterative process. But for individuals who already struggle with expressive language, executive functioning, or cognitive flexibility, the process itself may become overwhelming.

Ironically, a tool designed to reduce cognitive load may sometimes introduce a different kind of cognitive labour.

This experience has made me increasingly aware that AI accessibility is not simply about having access to the technology itself. It is also about whether the technology can adapt to differences in users’ cognitive profiles, including variations in communication, language processing, and executive functioning (Xu et al, 2025).

At present, much of the burden still falls on the user to adapt to the AI, rather than the AI adapting sufficiently to the user.

The Risk of Expanding Invisible Inequalities

There is another important concern.

As AI becomes normalized, society may begin assuming that everyone should automatically perform at a higher level simply because these tools now exist.

We have seen similar patterns emerge with smartphones, online systems, and digital platforms. Technologies intended to increase efficiency sometimes quietly increase expectations instead.

If AI helps many people work faster, educational and workplace systems may respond by:

• increasing productivity expectations

• accelerating timelines

• reducing accommodations

• assuming greater independence

• overlooking invisible cognitive effort

This matters because neurodivergence is not simply about access to tools. It is also about differences in cognitive load, processing style, communication, sensory experience, and adaptability.

Without careful attention to accessibility and inclusion, AI could unintentionally widen disparities between those who can rapidly adapt to emerging technologies and those who cannot (Melo-López et al., 2025).

Designing AI Around Human Diversity

Research examining inclusive human-AI interaction increasingly emphasizes the importance of adaptive and accessible design approaches for neurodivergent users (Xu et al., 2025).

Perhaps the most important question is not whether neurodivergent individuals can adapt to AI, but whether AI systems will adapt to diverse human needs.

There is a profound difference between:

• expecting humans to learn how to “speak AI,” and

• designing AI systems capable of understanding varied human communication styles.

The most accessible future may not depend on people mastering perfect prompts. Instead, it may depend on AI becoming:

• more conversational

• more intuitive

• more multimodal

• more personalized

• better at asking clarifying questions

• capable of adapting to different cognitive styles

Interestingly, many of the principles needed for accessible AI design already exist within neurodiversity advocacy and universal design frameworks (Melo-López et al., 2025; Zastudil et al., 2025):

• reducing hidden expectations

• providing explicit structure

• supporting flexible communication methods

• lowering cognitive friction

• accommodating different learning styles

• prioritizing accessibility from the outset

In that sense, neurodivergent perspectives may not simply matter in AI development, they may be essential to it.

Transitioning Into an AI-Shaped Future

Every major technological shift changes how humans interact with information, work, and one another. The transition into widespread AI use is no exception.

For neurodivergent individuals, this transition may bring:

• new opportunities for independence and expression

• powerful forms of cognitive support

• new accessibility barriers

• new expectations regarding productivity and communication

• new definitions of literacy and competence

As researchers, educators, clinicians, and communities, we should pay close attention not only to what AI can do, but also to:

• who benefits most,

• who struggles,

• what kinds of invisible labour these systems require, and

• whether emerging technologies truly accommodate human diversity.

Because the future of AI accessibility should not depend solely on whether neurodivergent individuals can adapt to technology.

It should also depend on whether technology is willing to adapt to them.

-- Anne TM Konkle, PhD

References

Dumitru, C., Abdulsahib, G. M., Khalaf, O. I., & Bennour, A. (2026). Integrating artificial intelligence in supporting students with disabilities in higher education: An integrative review. Technology and Disability. 38(1), 3-24. https://doi.org/10.1177/10554181251355428

Holmes, W., Bialik, M., & Fadel, C. (2019). Artificial Intelligence in Education: Promises and Implications for Teaching and Learning. Boston: Center for Curriculum Redesign.

Hwang, Y., Lee, J.H., Shin, D. (2023). What is prompt literacy? An exploratory study of language learners' development of new literacy skill using generative AI. arXiv.org Nov 9. DOI:10.48550/arXiv.2311.05373

Melo-López, V.-A., Basantes-Andrade, A., Gudiño-Mejía, C.-B., Hernández-Martínez, E. (2025). The impact of artificial intelligence on inclusive education: A systematic review. Education Sciences, 15(5), 539. https://doi.org/10.3390/educsci15050539

Ogut, B., Yin, M., Vu, H., Hicks, J., Circi, R. (2025). Universal by design: Unveiling the effectiveness of accommodations and universal design features through process data. Educational Measurement: Issues and Practice, 44(4): 18-32. https://doi.org/10.1111/emip.70007

Riefle, L., Hemmer, P., Benz, C., Vossing, M., Pries, J. (2022). On the influence of cognitive styles on users’ understanding of explanations. Forty-Third International Conference on Information Systems, Copenhagen.

Ronksley-Pavia, M., Nguyen, L., Wheeley, E., Rose, J., Neumann, M.M., Bigum, C., Neumann, D. L. (2025). A scoping literature review of generative artificial intelligence for supporting neurodivergent school students. Computers and Education: Artificial Intelligence, 9: 100437. https://doi.org/10.1016/j.caeai.2025.100437

Tenorio, M., Campos, R., Karmiloff-Smith, A. (2014). What standardized tests ignore when assessing individuals with neurodevelopmental disorders. Estudios de psicologia, 35 (2): 426-437. doi: 10.1080/02109395.2014.922264

Xu, Z., Liu, F., Xia, G., Duan, Y., Yu, L. (2025). A scoping review of inclusive and adaptive human-AI interaction design for neurodivergent users. Disability and rehabilitation: Assistive technology, 12: 1-19. doi: 10.1080/17483107.2025.2579822

Zastudil, C., Smith, D.H., Tohamy, Y., Nasimova, R., Montross, G., Macneil, S. (2025). Neurodiversity in Computing Education Research: A Systematic Literature Review. ITiCSE 2025: Proceedings of the 30th ACM Conference on Innovation and Technology in Computer Science Education, 1: 114 – 120. https://doi.org/10.1145/3724363.3729088