Adnan KhanAdnanProduct & AI
San Francisco Bay Area
Get in touch

Source note3 min read

The Context Access Divide: Interaction-Level Architecture as a Complementary Dimension of Agentic Inequality

Source

arXiv

arxiv.org/abs/2607.08495v1

Read the source

Source

Plain-Language Brief

The text introduces a new concept called the 'Context Access Divide' (CAD). This divide refers to the qualitative difference in AI utility experienced by two users who have similar access to AI agents, but one system can autonomously retrieve context from the user's knowledge corpus while the other requires manual input. The CAD is significant for knowledge-intensive workers as it determines whether the human or the AI system bears the cognitive burden of context curation. The text proposes 'contextuality', a new dimension of AI-mediated inequality, to complement the existing framework by Sharp et al. and discusses its implications for knowledge work stratification and AI platform governance.

What the Document Says

  • Sharp et al. (2025) introduced 'agentic inequality' as a framework for analyzing disparities in access to AI agents.
  • The framework by Sharp et al. has three dimensions: availability, quality, and quantity.
  • A new concept called the 'Context Access Divide' (CAD) is proposed, which addresses a finer level divide operating at the individual interaction level.
  • Two users with similar access to AI agents may have qualitatively different AI utility due to whether the system can autonomously retrieve context or requires manual input.
  • For knowledge-intensive workers, the CAD constitutes a significant threshold in AI usefulness, as it determines who bears the cognitive burden of context curation.
  • Contextuality is proposed as a new dimension of AI-mediated inequality that complements but is not reducible to the Sharp et al. framework.
  • The text formalizes the CAD with a probabilistic model grounded in the fan effect literature in cognitive psychology.
  • Manual context attachment leads to a combinatorial collapse in task-success probability as corpus size and task conjunctivity grow, while dynamic retrieval architectures are structurally insulated from this collapse.
  • The text discusses the technical basis of the divide in the Model Context Protocol (MCP) and retrieval-augmented generation (RAG) architectures.
  • The implications for knowledge work stratification and AI platform governance are examined.

Key Concepts Mentioned

  • agentic inequality
  • Context Access Divide (CAD)
  • contextuality
  • AI utility
  • knowledge-intensive workers
  • dynamic retrieval
  • Manual Attachment
  • fan effect
  • Model Context Protocol (MCP)
  • retrieval-augmented generation (RAG)

Claims Made by the Document

  • The Sharp et al. framework lacks a dimension addressing the individual interaction level.
  • The CAD is significant for knowledge-intensive workers as it determines who bears the cognitive burden of context curation.
  • Dynamic retrieval architectures are structurally insulated from combinatorial collapse in task-success probability compared to manual context attachment.

Evidence Used

  • The text refers to the work by Sharp et al. (2025) and explains their framework.
  • The CAD is explained through its impact on knowledge-intensive workers and its contrast with manual input.
  • The formalization of the CAD is described, including the probabilistic model and its grounding in cognitive psychology.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

Sharp et al. (2025) introduce 'agentic inequality' as a framework for analyzing disparities in access to AI agents across three dimensions: availability, quality, and quantity. These person- and organization-level dimensions characterize who can access agents and at what capability, but do not address a structurally important divide operating at a finer level: the individual interaction. Two users with nominally equivalent agent access may experience qualitatively different AI utility depending on whether the system can autonomously retrieve context from the user's knowledge corpus (Dynamic Context Retrieval) or requires the user to manually identify and attach relevant documents at each query (Manual Attachment). We term this the Context Access Divide (CAD). For knowledge-intensive workers whose intellectual capital spans tens of thousands of files, the CAD constitutes a qualitative threshold in AI usefulness: below it, the cognitive burden of context curation falls on the human, reproducing the inefficiencies AI is meant to eliminate. We propose contextuality -- the degree to which an AI system autonomously accesses a user's accumulated knowledge capital -- as a dimension of AI-mediated inequality that complements, but is not reducible to, the Sharp et al. framework. We formalize the CAD with a probabilistic model grounded in the fan effect literature in cognitive psychology, demonstrating that manual context attachment leads to a combinatorial collapse in task-success probability as corpus size and task conjunctivity grow, while dynamic retrieval architectures are structurally insulated from this collapse. We analyze the technical basis of this divide in the Model Context Protocol (MCP) and retrieval-augmented generation (RAG) architectures, and examine its implications for knowledge-work stratification and AI platform governance.