Museum Wire
Law 0 · Katie's LawEvery system is shaped by the human drive to do less work. This is not a flaw. It is the economic force that produces all software — and all software failure.◆Law I · Boundary CollapseWhen data crosses into a system that interprets structure, without being constrained, it becomes executable.◆◆ 2026 IncidentAxios. 70 Million Downloads a Week. North Korea Inside.◆Law II · Ambient AuthorityWhen a system trusts the presence of a credential instead of verifying the intent behind it, authentication becomes indistinguishable from authorization.◆◈ AXM-001Set Theory — Membership, Boundaries, and Belonging◆Law III · Transitive TrustWhen a system inherits trust from a source it did not verify, the attack surface extends to everything that source touches.◆◆ 2026 IncidentClaude Code — The Accept-Data-Loss Flag◆Law IV · Complexity AccretionSystems do not become complex. They accumulate complexity — one reasonable decision at a time — until no single person can hold the whole in their head.◆Law V · Temporal CouplingCode that assumes sequential execution, stable state, or consistent timing will fail the moment concurrency, scale, or latency proves the assumption wrong.◆◆ 2026 IncidentCopy Fail — 732 Bytes to Root on Every Linux Distribution◆◈ AXM-002Boolean & Propositional Logic — True, False, and the Excluded Middle◆Law VI · Observer InterferenceWhen the system that monitors health becomes a participant in the system it monitors, observation becomes a failure vector.◆◆ 2025Amazon Kiro — The 13-Hour Outage◆◆ 2025Operation Chrysalis: The Notepad++ Supply Chain Hijack◆◆ 2025Replit Agent — The Vibe Code Wipe◆◆ 2025Shai-Hulud — The npm Worm That Ate Its Own Ecosystem◆◆ 2024Air Canada Chatbot — The Policy That Wasn't◆◆ 2024Change Healthcare — One-Third of US Healthcare, One Missing MFA◆◆ 2024CrowdStrike — The Security Update That Broke the World◆◆ 2024Google Gemini Image Generation — The Six-Day Pause◆◆ 2024XZ Utils — The Two-Year Infiltration◆◆ 20233CX — The Supply Chain That Ate Another Supply Chain◆◆ 2023Amazon Prime Video — The Per-Frame State Machine◆◆ 2023Bing Sydney — The Chatbot That Went Rogue◆◆ 2023Samsung ChatGPT Leak — The Employee Who Pasted the Secret◆EFFODE · LEGE · INTELLEGE◆Law 0 · Katie's LawEvery system is shaped by the human drive to do less work. This is not a flaw. It is the economic force that produces all software — and all software failure.◆Law I · Boundary CollapseWhen data crosses into a system that interprets structure, without being constrained, it becomes executable.◆◆ 2026 IncidentAxios. 70 Million Downloads a Week. North Korea Inside.◆Law II · Ambient AuthorityWhen a system trusts the presence of a credential instead of verifying the intent behind it, authentication becomes indistinguishable from authorization.◆◈ AXM-001Set Theory — Membership, Boundaries, and Belonging◆Law III · Transitive TrustWhen a system inherits trust from a source it did not verify, the attack surface extends to everything that source touches.◆◆ 2026 IncidentClaude Code — The Accept-Data-Loss Flag◆Law IV · Complexity AccretionSystems do not become complex. They accumulate complexity — one reasonable decision at a time — until no single person can hold the whole in their head.◆Law V · Temporal CouplingCode that assumes sequential execution, stable state, or consistent timing will fail the moment concurrency, scale, or latency proves the assumption wrong.◆◆ 2026 IncidentCopy Fail — 732 Bytes to Root on Every Linux Distribution◆◈ AXM-002Boolean & Propositional Logic — True, False, and the Excluded Middle◆Law VI · Observer InterferenceWhen the system that monitors health becomes a participant in the system it monitors, observation becomes a failure vector.◆◆ 2025Amazon Kiro — The 13-Hour Outage◆◆ 2025Operation Chrysalis: The Notepad++ Supply Chain Hijack◆◆ 2025Replit Agent — The Vibe Code Wipe◆◆ 2025Shai-Hulud — The npm Worm That Ate Its Own Ecosystem◆◆ 2024Air Canada Chatbot — The Policy That Wasn't◆◆ 2024Change Healthcare — One-Third of US Healthcare, One Missing MFA◆◆ 2024CrowdStrike — The Security Update That Broke the World◆◆ 2024Google Gemini Image Generation — The Six-Day Pause◆◆ 2024XZ Utils — The Two-Year Infiltration◆◆ 20233CX — The Supply Chain That Ate Another Supply Chain◆◆ 2023Amazon Prime Video — The Per-Frame State Machine◆◆ 2023Bing Sydney — The Chatbot That Went Rogue◆◆ 2023Samsung ChatGPT Leak — The Employee Who Pasted the Secret◆EFFODE · LEGE · INTELLEGE◆
Keyboard Navigation
W↑
A←
S↓
D→
or arrow keys · M for map · Q to exit
← Back to Incident Room
2023data losscorporate

Samsung ChatGPT Leak — The Employee Who Pasted the Secret

Three Samsung semiconductor employees pasted proprietary chip design source code, internal meeting notes, and confidential test results into ChatGPT. The data was permanently incorporated into OpenAI's training data. Samsung subsequently banned ChatGPT for internal use.

3 min read
Root Cause

Employees treated ChatGPT as a private productivity tool — an extension of their private cognitive workspace. ChatGPT is a public service. All inputs are potentially retained for model training. The employees understood what they were pasting. They did not understand where they were pasting it.

Aftermath

Samsung banned generative AI tools for internal use. Multiple large corporations issued similar bans (Apple, Amazon, JPMorgan). Accelerated enterprise AI governance discussions. The data cannot be retrieved.

The Incident

In April 2023, Bloomberg and The Economist Korea reported that three Samsung semiconductor employees had entered sensitive proprietary information into ChatGPT within a 20-day period:

1. An engineer pasted proprietary semiconductor source code into ChatGPT and asked it to check for bugs.

2. A second employee pasted source code from the same database, asking ChatGPT to optimize it.

3. A third employee transcribed a confidential internal meeting — including notes from a business strategy session — and asked ChatGPT to summarize it.

In all three cases, the data was entered into OpenAI's service, which at the time retained user inputs for model improvement. The proprietary information — chip designs representing years of competitive advantage — was permanently incorporated into a training corpus accessible to OpenAI.

Samsung banned ChatGPT for internal use within weeks, as did dozens of major corporations.

The Pattern

This is Ambient Authority (Law II) operating at the employee layer. The employees had legitimate authority over the data in the context where they worked with it — their workstations, their internal tools, their domain. They did not have authority to disclose it externally. ChatGPT looked like an internal tool because it behaved like one (it felt private, it felt like a personal assistant) but it was a public endpoint.

The mental model failure: "I am having a private conversation with an AI assistant." The operational reality: "I am submitting text to a third-party's training pipeline."

This is the same mental model failure behind every CSRF attack: the browser makes a request that looks like it comes from the user's session, because it does, but the intent behind the request is not the user's. The employee pasted proprietary code because the interface felt like a private workspace. The interface was not private.

The Exhibit

This incident maps to [The Boundary Collapse](/exhibits/boundary-collapse) pattern (Law I): the boundary between proprietary internal data and a public external service was not enforced by technology, only by employee understanding. The employees understood what they were pasting. They did not understand the boundary they were crossing.

This is the distinction the museum keeps trying to surface: you cannot defend a boundary that is not implemented as a constraint. "Employees shouldn't do this" is a policy. "Employees can't do this" is a control. Samsung had a policy. They did not have a control.

Why It Matters

The Samsung incident was not an edge case. Within three months, Amazon internal Slack channels leaked that Amazon employees were pasting AWS confidential information into ChatGPT. Apple issued a ban. JPMorgan issued a ban. The same incident pattern replicated across industries — because the mental model failure (private conversation vs. public endpoint) is universal, not specific to Samsung.

The data didn't leak. It was pasted. The distinction matters because the fix is different: you can't patch a paste.
Techniques
data exfiltration via llmambient authorityinsider data exposure