When AI Reads the File

An Apple Workers’ Comp Record Where the Documents and the Denials Don’t Match

A 14-year workers’ compensation record involving Apple and Sedgwick analyzed as a single archive using modern AI tools.

• 14 years of records preserved

• Medical requests compared directly to denial reasoning

• AI-assisted review of the full file

Escalation Timeline

How concerns regarding the claim were reported internally at Apple

Over the course of more than a decade, concerns regarding the handling of the claim were reported through multiple internal channels at Apple.

2012

Apple Store workplace injury.

2012–2016

Concerns regarding claim administration and treatment denials reported through Apple Human Resources channels.

2016

Concerns regarding Sedgwick claim handling reported to Apple’s Business Conduct Team.

Leadership Escalation

November 2017

Whistleblower report sent to Apple Senior Vice President Deirdre O’Brien regarding claim administration and treatment denials.

August 2020

Whistleblower video presentation sent to Apple CEO Tim Cook describing concerns regarding claim handling and medical treatment denials.

July 2024

Email sent to Tim Cook raising concerns regarding the accuracy of a Medicare Set-Aside report prepared for the claim.

February 2025

Whistleblower report and AI-assisted analysis of claim records sent to Tim Cook and Apple General Counsel Kate Adams.

2026

AI-assisted analysis of the records examined in this project conducted using ChatGPT, Claude, Grok, and Gemini.

What the Timeline Shows

This timeline is not presented to suggest that a single email or report should have resolved every issue immediately.

It is presented to show something narrower and more concrete:

the concerns were raised repeatedly, through multiple internal channels, over a period of years.

What began as a workplace injury and claim file eventually became something else:

a documented record of internal reporting, leadership escalation, and repeated efforts to identify discrepancies between the written record and the reasoning used to deny treatment.

That is part of what makes this project different from ordinary online complaints about health-care denials.

The issue is not simply that treatment was denied.

The records examined in this project allow readers to compare two things:

What the Documents Say

vs.

What Decision-Makers Said the Documents Meant.

This project uses modern language-model tools to analyze the records examined here and highlight where explanations diverge from the written record.