Skip to main content

AI Displacement Index Methodology

How LayoffSignal calculates workforce automation risk.

01 What is the AI Displacement Index?

The AI Displacement Index is a proprietary metric (0-100) that evaluates the likelihood that a recently executed layoff was driven—directly or indirectly—by advances in generative AI, automation, and operational restructuring. A high score indicates that the roles eliminated overlap heavily with tasks modern Large Language Models (LLMs) and agents are highly capable of performing.

02 Data Sourcing & NLP Pipeline

We do not rely on PR spin. LayoffSignal processes ground-truth data from two primary avenues:

  • Statutory Disclosures: Parsing exact job titles and headcount reductions submitted to state governments via WARN Act notices.
  • Corporate Filings: Analyzing SEC 8-K filings and official restructuring announcements to extract the real roles affected (e.g., "Customer Support", "QA", "Copywriting", "Middle Management").

03 LLM Scoring Engine

We pass the extracted roles and company context into our AI scoring pipeline. The pipeline evaluates the layoff event against several automation criteria:

High Risk (75-100)

Routine cognitive tasks, customer support, data entry, junior copywriting, and basic QA. Highly susceptible to immediate LLM substitution.

Moderate Risk (40-74)

Middle management, software engineering (junior/mid), financial analysis. Roles where AI provides massive leverage, reducing total headcount needs.

Low Risk (0-39)

Physical operations, hardware engineering, field sales, complex specialized compliance. Roles requiring physical presence or highly nuanced human relationship building.

04 Talent Arbitrage

Alongside the Displacement Index, we calculate the Talent Arbitrage Index. This tracks whether a company is laying off expensive US-based talent while concurrently shifting those exact reqs to offshore markets (like India) or automating them entirely to improve Gross Margins and free cash flow.

Proprietary Methodology · The scoring models, analytical frameworks, and data processing pipelines described on this page are proprietary to LayoffSignal. This overview is provided for transparency and evaluation purposes only.

The specific implementation details — including model weights, prompt engineering, scoring algorithms, data enrichment logic, and pipeline architecture — are trade secrets of LayoffSignal and are not disclosed in this document.

© 2026 LayoffSignal. All rights reserved. Reproduction, redistribution, or derivative use of the methodologies described herein without express written permission is prohibited.