pith. sign in

arxiv: 2604.25455 · v1 · submitted 2026-04-28 · 💻 cs.HC · cs.AI· cs.CY

Generative UI as an Accessibility Bridge: Lessons from C2C E-Commerce

Bektur Ryskeldiev This is my paper

Pith reviewed 2026-05-07 15:43 UTC · model grok-4.3

classification 💻 cs.HC cs.AIcs.CY
keywords generative UIweb accessibilityC2C e-commerceuser-generated contentscreen readersblind usersolder adultsHCI
0
0 comments X

The pith

Generative UI can create real-time adapted interfaces that overcome accessibility barriers in C2C e-commerce platforms where static standards fail.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

In customer-to-customer marketplaces, user-generated content like photos and descriptions varies too much for fixed accessibility rules to cover every case. The paper draws on six studies with blind, low-vision, and older adult users to show that generative UI, which builds interfaces at the point of use, can adapt to these specifics. Concrete examples include regenerating HTML to improve screen reader support, using conversation to guide older sellers, and giving audio directions to help blind sellers frame photos properly. If correct, this would mean accessibility improvements no longer depend only on preemptive design or compliance but can happen dynamically for each user and listing.

Core claim

The central claim is that generative UI produces adapted interfaces at the point of use to address barriers that static design cannot anticipate in C2C e-commerce. Evidence comes from three interventions tested in studies: HTML regeneration for screen readers, conversational guidance for older sellers, and audio-guided photo framing for blind sellers. These demonstrate how runtime generation bridges gaps left by standards, leading to the view that generative UI extends beyond the screen, complements ability-based design, and changes the designer's role to specifying policies.

What carries the argument

Runtime generative UI that takes user-generated content and user needs as input to produce tailored interfaces on the fly.

If this is right

  • Generative UI can fix issues from unpredictable content such as off-frame photos or incomplete descriptions.
  • It supports adaptations in multiple modalities including audio and conversational interfaces.
  • Designers shift focus from creating fixed layouts to defining rules and policies for generation.
  • The method works together with rather than instead of traditional ability-based design approaches.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • These ideas could extend to other sites with highly variable user content like social networks or review platforms.
  • Real-world deployment would require testing how often the generation introduces new usability errors.
  • Policies for generation might need user-controlled options to maintain trust and personalization.
  • Integration with existing accessibility tools could amplify the benefits across more user groups.

Load-bearing premise

Generative systems can reliably interpret the varied user-generated content and specific user needs to produce accurate and helpful adaptations without creating new errors or trust issues.

What would settle it

A follow-up experiment measuring completion rates and satisfaction for the three interventions against standard non-generative interfaces, to check if adaptations succeed more often than they fail or confuse users.

Figures

Figures reproduced from arXiv: 2604.25455 by Bektur Ryskeldiev.

Figure 1
Figure 1. Figure 1: Generative UI as an accessibility bridge in C2C e-commerce. Inaccessible user-generated content (left) is transformed through view at source ↗
read the original abstract

Web accessibility rests on static standards and developer compliance. That model frays in platforms where content is user-generated: photos arrive blurry or off-frame, descriptions skip size and condition, and page structure shifts from listing to listing. Drawing on six studies conducted between 2022 and 2025 with blind, low-vision, and older adult users of customer-to-customer (C2C) marketplaces, I argue that generative UI can produce adapted interfaces at the point of use, addressing barriers that static design cannot anticipate. Three interventions from this program -- HTML regeneration for screen readers, conversational guidance for older sellers, and audio-guided photo framing for blind sellers -- demonstrate how runtime generation can bridge gaps that standards leave open. I outline what these findings imply for HCI practice: generative UI extends beyond the screen, complements rather than replaces ability-based design, and shifts the designer's role from specifying layouts to specifying policies. This is an expanded arXiv version of a position paper accepted at the CHI 2026 workshop "What does Generative UI mean for HCI Practice?"

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 1 minor

Summary. The paper claims that generative UI can produce adapted interfaces at the point of use to address accessibility barriers in C2C e-commerce platforms that static standards cannot anticipate, particularly with user-generated content. Drawing on six studies (2022-2025) with blind, low-vision, and older adult users, it presents three interventions—HTML regeneration for screen readers, conversational guidance for older sellers, and audio-guided photo framing for blind sellers—as existence proofs. The paper outlines implications for HCI practice: generative UI extends beyond the screen, complements ability-based design, and shifts the designer's role from specifying layouts to specifying policies. It is an expanded arXiv version of a position paper accepted at the CHI 2026 workshop on Generative UI.

Significance. If the lessons hold, the work contributes to HCI by demonstrating how generative technologies can extend accessibility support into dynamic, user-generated environments where traditional compliance-based approaches are insufficient. It offers concrete examples and practice-oriented implications that could guide future research on runtime adaptation and inclusive design for platforms like marketplaces.

major comments (1)
  1. Abstract: The central claim rests on six user studies supporting the three interventions, yet no methodological details, participant numbers, study designs, data analysis methods, or limitations are provided. This is load-bearing because the interventions are presented as demonstrations drawn directly from these studies, and without such information the strength of evidence for runtime generation bridging gaps cannot be evaluated.
minor comments (1)
  1. The implications section could clarify how 'specifying policies' differs in practice from traditional UI specification, with at least one concrete example tied to one of the three interventions.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful review and for identifying the need to clarify the evidential foundation of the position paper. We address the single major comment below.

read point-by-point responses

  1. Referee: Abstract: The central claim rests on six user studies supporting the three interventions, yet no methodological details, participant numbers, study designs, data analysis methods, or limitations are provided. This is load-bearing because the interventions are presented as demonstrations drawn directly from these studies, and without such information the strength of evidence for runtime generation bridging gaps cannot be evaluated.

    Authors: We agree that the abstract itself contains no methodological details. This manuscript is an expanded position paper (originating from a CHI 2026 workshop acceptance) whose contribution is the synthesis of practice implications rather than the presentation of new empirical results. The six studies (2022–2025) are our previously published empirical work on C2C accessibility barriers; they are cited throughout the body of the paper, where the interventions are explicitly framed as “existence proofs” drawn from that program. Full participant numbers, designs, analysis methods, and limitations appear in those cited publications. We will revise the abstract to add one sentence stating that the studies belong to our prior research program and that methodological details are reported in the referenced works. This change preserves the position-paper character while allowing readers to locate the primary evidence. revision: yes

Circularity Check

0 steps flagged

No significant circularity: argument grounded in external user studies

full rationale

The paper is a position paper that summarizes lessons from six independent user studies (2022-2025) with blind, low-vision, and older adult C2C marketplace users. Its central claim—that three generative interventions (HTML regeneration, conversational guidance, audio-guided framing) demonstrate runtime adaptation for accessibility gaps—is presented as an existence proof drawn from those empirical results rather than any internal derivation, fitted parameter, or self-referential definition. No equations, ansatzes, uniqueness theorems, or self-citations appear as load-bearing steps in the provided text. The argument is self-contained against the external benchmarks of the cited studies and does not reduce any prediction or conclusion to its own inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

This is a qualitative position paper grounded in user studies. No free parameters or invented entities are introduced. The work relies on standard HCI assumptions about the validity of user-centered research.

axioms (1)
  • domain assumption Findings from user studies with disabled and older adult participants provide reliable guidance for accessibility interventions.
    The central argument is built directly on the outcomes of the six described studies.

pith-pipeline@v0.9.0 · 5484 in / 1370 out tokens · 53855 ms · 2026-05-07T15:43:25.581643+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

17 extracted references · 14 canonical work pages

  1. [1]

    Krzysztof Gajos and Daniel S. Weld. 2004. SUPPLE: Automatically Generating User Interfaces. InProceedings of the 9th International Conference on Intelligent User Interfaces (IUI ’04). ACM, 93–100. doi:10.1145/964442.964461

    work page doi:10.1145/964442.964461 2004

  2. [2]

    Gajos, Jacob O

    Krzysztof Z. Gajos, Jacob O. Wobbrock, and Daniel S. Weld. 2007. Automatically Generating User Interfaces Adapted to Users’ Motor and Vision Capabilities. InProceedings of the 20th Annual ACM Symposium on User Interface Software and Technology (UIST ’07). ACM, 231–240. doi:10.1145/1294211.1294253

    work page doi:10.1145/1294211.1294253 2007

  3. [3]

    Kunal Gupta, Ken Takaki, Ayaka Tsutsui, Kentaro Teramoto, and Bektur Ryskeldiev. 2025. AccessibleShip: Exploring Packaging and Shipping Barriers for Visually Impaired Users in C2C E-Commerce. InExtended Abstracts of the CHI Conference on Human Factors in Computing Systems (CHI EA ’25). ACM. doi:10.1145/3706599.3706665

    work page doi:10.1145/3706599.3706665 2025

  4. [4]

    Churchill, Abigail Sellen, Jeffrey Nichols, and David R

    Siân Lindley, Jack Williams, Yining Cao, Haijun Xia, Elizabeth F. Churchill, Abigail Sellen, Jeffrey Nichols, and David R. Karger. 2026. What does Generative UI mean for HCI Practice?. InExtended Abstracts of the CHI Conference on Human Factors in Computing Systems (CHI EA ’26). ACM

    2026

  5. [5]

    I Bought This for Me to Look More Ordinary

    Guanhong Liu, Xianghua Ding, Chun Yu, Lan Gao, Xingyu Chi, and Yuanchun Shi. 2019. “I Bought This for Me to Look More Ordinary”: A Study of Blind People Doing Online Shopping. InProceedings of the 2019 CHI Conference on Human Factors in Computing Systems (CHI ’19). ACM, 1–11. doi:10.1145/3290605.3300602

    work page doi:10.1145/3290605.3300602 2019

  6. [6]

    Yoshiki Nagatani, Kazuki Takazawa, Masaaki Sadasue, Mariko Kobayashi, Kentaro Teramoto, Nozomu Yoshida, Kotaro Oomori, Akihisa Shitara, Koyomi Kobayashi, Aiko Kuroda, and Misato Takahashi. 2022. How Visually Impaired People Use Information Media and E-Commerce in Japan. In HCI International 2022 Posters, Constantine Stephanidis, Margherita Antona, and Sta...

    work page doi:10.1007/978-3-031-06394-7_70 2022

  7. [7]

    OpenAI. 2024. GPT-4o. https://platform.openai.com/docs/models/gpt-4o

    2024

  8. [8]

    Subin Park, Hyunggu Jung, Bektur Ryskeldiev, Jihyun Lee, Gwanhee Lee, Hyeonhak Jeong, Minki Chun, and Kentaro Teramoto. 2024. Toward Closed-Domain Conversational Item Listing Assistant for Improvement of Experiences of Older Adults in Customer-to-Customer (C2C) Marketplaces. InProceedings of the 26th International ACM SIGACCESS Conference on Computers and...

    work page doi:10.1145/3663548.3688522 2024

  9. [9]

    Bektur Ryskeldiev, Kotaro Hara, Mariko Kobayashi, and Koki Kusano. 2022. Investigating Accessibility Challenges and Opportunities for Users with Low Vision Disabilities in Customer-to-Customer (C2C) Marketplaces. InProceedings of the 24th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS ’22). ACM. doi:10.1145/3517428.3550390

    work page doi:10.1145/3517428.3550390 2022

  10. [10]

    Bektur Ryskeldiev, Mariko Kobayashi, Kentaro Teramoto, and Koki Kusano. 2023. Towards Immersive Inclusivity for C2C: How Immersive Multimodal Interactions Can Make Online Customer-to-Customer Shopping More Inclusive. InProceedings of the Asian CHI Symposium (AsianHCI ’23). ACM, 84–88. doi:10.1145/3604571.3604586

    work page doi:10.1145/3604571.3604586 2023

  11. [11]

    Madaio, Luke Stark, Jennifer Wortman Vaughan, and Hanna Wallach

    Abigale Stangl, Meredith Ringel Morris, and Danna Gurari. 2020. “Person, Shoes, Tree. Is the Person Naked?” What People with Vision Impairments Want in Image Descriptions. InProceedings of the 2020 CHI Conference on Human Factors in Computing Systems (CHI ’20). ACM, 1–13. doi:10.1145/ 3313831.3376404

    work page arXiv 2020

  12. [12]

    Stangl, Esha Kothari, Suyog D

    Abigale J. Stangl, Esha Kothari, Suyog D. Jain, Tom Yeh, Kristen Grauman, and Danna Gurari. 2018. BrowseWithMe: An Online Clothes Shopping Assistant for People with Visual Impairments. InProceedings of the 20th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS ’18). ACM, 107–118. doi:10.1145/3234695.3236337

    work page doi:10.1145/3234695.3236337 2018

  13. [13]

    I Am a Blind Seller!

    Ayaka Tsutsui, Kengo Tanaka, Yaman Yu, and Bektur Ryskeldiev. 2025. “I Am a Blind Seller!”: Picture Taking Assistance for Visually Impaired Individuals for Participation as Sellers in Customer to Customer (C2C) Marketplaces. InExtended Abstracts of the CHI Conference on Human Factors in Computing Systems (CHI EA ’25). ACM. doi:10.1145/3706599.3720150

    work page doi:10.1145/3706599.3720150 2025

  14. [14]

    W3C. 2018. Web Content Accessibility Guidelines (WCAG) 2.1. (2018). https://www.w3.org/TR/WCAG21/

    2018

  15. [15]

    Ruolin Wang, Zixuan Chen, Mingrui Ray Zhang, Zhaoheng Li, Zhixiu Liu, Zihan Dang, Chun Yu, and Xiang ’Anthony’ Chen. 2021. Revamp: Enhancing Accessible Information Seeking Experience of Online Shopping for Blind or Low Vision Users. InProceedings of the 2021 CHI Conference on Human Factors in Computing Systems (CHI ’21). ACM, Article 494, 14 pages. doi:10...

    work page doi:10.1145/3411764.3445547 2021

  16. [16]

    Wobbrock, Shaun K

    Jacob O. Wobbrock, Shaun K. Kane, Krzysztof Z. Gajos, Susumu Harada, and Jon Froehlich. 2011. Ability-Based Design: Concept, Principles and Examples.ACM Transactions on Accessible Computing3, 3 (2011), 1–27. doi:10.1145/1952383.1952384

    work page doi:10.1145/1952383.1952384 2011

  17. [17]

    Yaman Yu, Bektur Ryskeldiev, Ayaka Tsutsui, Matthew Gillingham, and Yang Wang. 2025. From Cluttered to Clear: Improving the Web Accessibility Design for Screen Reader Users in E-commerce With Generative AI. InProceedings of the 27th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS ’25). ACM. doi:10.1145/3663547.3746353

    work page doi:10.1145/3663547.3746353 2025