Adnan KhanAdnanProduct & AI
San Francisco Bay Area
Get in touch

Source note3 min read

Structural Silence: When AI Infrastructure Fails Speakers of Underrepresented Languages

Source

arXiv

arxiv.org/abs/2608.12278v1

Read the source

Source

  • Source name: arXiv
  • Source kind: paper
  • Authors / organization: Avijit Roy, Proma Roy
  • Venue: arXiv: cs.CL
  • Published: 2026-08-12
  • Collected: 2026-08-13
  • URL: http://arxiv.org/abs/2608.12278v1
  • PDF: https://arxiv.org/pdf/2608.12278v1
  • DOI: UNKNOWN
  • Registry group: academic_apis_and_preprints
  • Evidence role: canonical_academic_index
  • Language: en
  • Country / region: UNKNOWN

Plain-Language Brief

The paper discusses how AI tools for education and language support can disadvantage speakers of underrepresented languages before training, using Bengali as a case study. The authors identify four main issues: a web presence gap, data deficit in multilingual corpora, tokenization penalty due to Bengali's alphasyllabary script, and connectivity exclusion. They argue that these are not just technical limitations but structural barriers, and suggest offline-first design as an equity-oriented strategy.

What the Document Says

  • AI tools for education and language support can disadvantage speakers of underrepresented languages.
  • Bengali accounts for less than 0.5% of global web content despite representing nearly 4% of the global population.
  • There is a 67:1 training-token deficit between English and Bengali in major multilingual corpora.
  • Bengali's alphasyllabary script has a tokenization penalty that compounds the data deficit through higher token fertility.
  • Individual internet penetration is 36.5% in rural areas compared with 71.4% in urban areas for Bengali speakers.
  • These issues reflect historical resource-allocation decisions, institutional priorities, and design defaults that did not center underrepresented languages in mainstream AI development.
  • The authors argue that dataset scarcity should be understood as a structural barrier rather than an isolated technical limitation.

Key Concepts Mentioned

  • Artificial Intelligence
  • Underrepresented Languages
  • Bengali
  • Education
  • Language Support
  • Web Content
  • Multilingual Corpora
  • Tokenization
  • Script
  • Connectivity

Claims Made by the Document

  • AI tools can disadvantage speakers of underrepresented languages.
  • Dataset scarcity is a structural barrier, not just a technical limitation.

Evidence Used

  • Less than 0.5% of global web content is in Bengali despite representing nearly 4% of the global population.
  • There is a 67:1 training-token deficit between English and Bengali in major multilingual corpora.
  • Bengali's alphasyllabary script has a tokenization penalty that compounds the data deficit through higher token fertility.
  • Individual internet penetration is 36.5% in rural areas compared with 71.4% in urban areas for Bengali speakers.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

  • What solutions do the authors propose to address these structural barriers in AI development for underrepresented languages?
  • How can offline-first design be implemented effectively as an equity-oriented infrastructure strategy?

Notes for Future Q&A

May include interpretation. Labeled where it does.

  • The paper focuses on Bengali as a case study but its findings may be applicable to other underrepresented languages.
  • The authors suggest offline-first design as an equity-oriented infrastructure strategy, but the details of this strategy are not discussed in the provided text.

Raw Abstract / Excerpt

Artificial intelligence tools for education and language support are increasingly framed as scalable responses to access gaps in under-resourced communities. Yet the infrastructure underlying these tools, including training corpora, tokenization schemes, evaluation benchmarks, and deployment architectures, can systematically disadvantage speakers of underrepresented languages before a model is trained. This paper examines these structural barriers through Bengali, one of the world's most widely spoken languages, focusing on AI-assisted education in low-connectivity environments. We identify four interlocking failures: a severe web presence gap, with Bengali accounting for less than 0.5% of global web content despite representing nearly 4% of the global population; a 67:1 training-token deficit between English and Bengali in major multilingual corpora; a tokenization penalty associated with Bengali's alphasyllabary script that compounds the data deficit through higher token fertility; and connectivity exclusion, with individual internet penetration at 36.5% in rural areas compared with 71.4% in urban areas. These failures reflect longstanding resource-allocation decisions, institutional priorities, and design defaults that did not center underrepresented languages in mainstream AI development. We argue that dataset scarcity should be understood as a structural barrier rather than an isolated technical limitation, and that offline-first design should be treated as an equity-oriented infrastructure strategy. We conclude with directions for linguistics and AI research aimed at reducing these structural inequalities.