The Index · Manufacturing

India AI-Native Index · Edition one · Public evidence only

JSW Steel

Scored on public evidence published between February 2025 and August 2026, reviewed through 30 August 2026, under the Edition 1 locked rule set (ANS-100/P v2.1, 1-floor on every question). Provisional result and evidence annexure sent to the company fourteen days before publication.

48.9Score out of 100
AI IntegratedBand · 41–60
BConfidence · 5 of 8 evidence categories
SecurityCeilings triggered

Bands are calibrated thresholds on the published scale, not ranks.

What this assessment is, and is not

The Index measures what a company has placed on the public record about its use of AI, within the evidence window, against a published rubric. It does not measure business performance, product quality, safety, customer outcomes, or the adequacy of internal controls. A low score on any question means that, in the public sources listed below, reviewed through the cut-off date, disclosure sufficient to satisfy the published criterion was not located. The company may maintain relevant practices or controls that are not publicly disclosed.

By dimension

Knowledge & context
50%
Agents & automation
62%
Model & infrastructure
31%
Data & AI security
33%
Organisation
44%
Economics & outcomes
75%
Business-model exposure
38%

Each dimension is the share of available points earned on the questions that could be scored from public evidence. 24 of 35 questions were scoreable; the remainder are outside the public-evidence instrument.

Evidence located

Questions on which the reviewed public sources contained disclosure meeting at least the level-2 anchor of the published rubric:

Level 3 — instances, or a disclosed number linked to a named system

  • production AI autonomy
  • AI deployed across functions
  • orchestration or integration platform
  • AI leadership and organisation
  • AI attribution to reported economics
  • AI-attributed operational rate
  • improvement trail and versioning

Level 2 — disclosed; not quantified or independently verified

  • named AI platform or knowledge substrate
  • internal AI tooling in use
  • hosting or infrastructure posture
  • canonical data platform
  • AI-specific investment
  • position on agentic or AI-mediated exposure
  • leadership commitment and strategic response

Level descriptions are those of the published rubric. The evidence annexure sent to the company lists the specific documents and passages relied on for each question.

Where disclosure meeting the criterion was not located

In the public sources listed below, reviewed through 30 August 2026, we did not locate disclosure sufficient to satisfy the Index's published criterion for the following. The company may maintain relevant practices or controls that are not publicly disclosed.

  • guardrail and evaluation documentation
  • model strategy or vendor architecture
  • open or external build surface
  • evaluation and benchmarking practice
  • API or integration surface
  • AI governance posture
  • AI talent or training figures
  • role redesign around AI
  • AI-native monetisation
  • product that could not exist before AI, with volumes

Ceiling

Security ceiling

Security ceiling triggered under the published rubric: the reviewed public evidence met the level-3 autonomy threshold, while public governance evidence remained below level 2. The resulting cap did not change the company's score. This does not establish the absence or adequacy of internal controls.

Sources reviewed

FY26 Integrated Annual Report (company-primary), Q2 and Q4/FY26 results presentations, FY26 operational update, FE CIO feature quoting company executive (Jul 2025), jswsteel.in digital-transformation disclosures; Project Athena material from 2019 admitted as background only, outside the window

Archive references

SourceOriginalArchiveCaptured
Transform — Integrated Report 2024–25originalarchived copy2026-09-02

Exchange filings and reports relied on:

All sources archived at scoring date; the complete source archive register is available on request via corrections@thecompounding.company. Short quotations, where used, are attributed and used for the purpose of review. Right-of-reply notice sent 10 September 2026 to the Company Secretary, Investor Relations, registered office, corporate communications and legal function. Company response: none received as at publication. No response is characterised as agreement or endorsement.

This assessment is based exclusively on public information available during the stated evidence window. It does not assess undisclosed internal systems or practices. Scores and bands are opinions produced by applying the published rubric to the cited factual record. Not investment advice. The Index is not affiliated with, endorsed by, or sponsored by any assessed company.

Corrections: material factual errors are corrected promptly and logged. See the corrections log and the methodology, including its removal policy.