TRV-2026-1995Certified recordPeer-reviewed

Generative AI and Social Inequality in Education: A Human-Centered Social Science Review of Global and Indonesian Evidence

This review examined how generative artificial intelligence (AI) can widen educational opportunity while also reproducing social inequality. An integrative review synthesized 24 sources: 17 Indonesian publications (2022–2026), identified through national journal portals and Google Scholar, and seven international policy documents and scholarly works (2016–2023) selected as conceptual anchors. Thematic analysis produced four themes: unequal access, unequal digital and AI literacy, unequal capacity to benefit from…

Education · Problem Space — documented harm · certified 2026-10-20 · v1 · article view · machine-readable

Current reading — problem

The benefits of generative AI in education depend on connectivity, devices, language resources, teacher competence, and evaluation skills, so unequal access and literacy can reproduce social inequality and raise governance risks.

What this doesn’t fix

The Indonesian evidence is mostly cross-sectional, self-reported, or from small single-institution samples with little evidence on learning outcomes, and the proposed model is conceptual rather than tested.

Evidence

Reader signal

How should this claim be treated?

Cite this record

Truvace Impact Record TRV-2026-1995, v1: “Generative AI and Social Inequality in Education: A Human-Centered Social Science Review of Global and Indonesian Evidence.” Truvace, 2026-10-20. /record/TRV-2026-1995 (accessed at citation time). sha256 95ff899c3fe9367a…

Calibration history

Every change to this record since certification, in the open. None yet — the reading has held since it entered the record.

  1. Certifiedv195ff899c3fe9…

    Certified into the record

Verify this record
How to verify without trusting this page

Fetch the canonical text of any version from /api/record/TRV-2026-1995 and hash it yourself — for example shasum -a 256 on the saved canonical field. The result must equal content_hash, and each version’s text ends with prev:followed by the prior version’s hash (version 1 chains to 64 zeros). If a single character of any version had been altered since certification, the chain would not reproduce.