---
title: "OpenAI’s new ChatGPT feature logs your keystrokes and stores them in plain text | SpinGraph: Safety framing"
description: "SpinGraph analysis of Google News: OpenAI's OpenAI’s new ChatGPT feature logs your keystrokes and stores them in plain text story: safety framing, The Shield, …"
	canonical: "https://stuffthatspins.com/spin/openais-new-chatgpt-feature-logs-your-keystrokes-and-stores-them-in-plain-text-the-next-web"
html: "https://stuffthatspins.com/spin/openais-new-chatgpt-feature-logs-your-keystrokes-and-stores-them-in-plain-text-the-next-web"
json: "https://stuffthatspins.com/spin/openais-new-chatgpt-feature-logs-your-keystrokes-and-stores-them-in-plain-text-the-next-web.json"
markdown: "https://stuffthatspins.com/spin/openais-new-chatgpt-feature-logs-your-keystrokes-and-stores-them-in-plain-text-the-next-web.md"
keywords: ["keystroke logging", "plain text storage", "ChatGPT privacy", "The Shield", "narrative intelligence"]
date: "2026-08-16T15:36:50+00:00"
modified: "2026-08-16T19:09:31.618069+00:00"
json_ld: |
  {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://stuffthatspins.com/#organization","name":"Stuff That Spins","url":"https://stuffthatspins.com/","description":"Know the moment AI knows your story. Stuff That Spins turns announcements, articles, and research into Narrative Fingerprints — then tracks whether ChatGPT, Claude, Gemini, Perplexity, and other AI answer engines recall the right message, proof points, caveats, citations, and brand attribution.","logo":{"@type":"ImageObject","url":"https://stuffthatspins.com/images/logo.png"},"sameAs":[]},{"@type":"NewsArticle","@id":"https://stuffthatspins.com/spin/openais-new-chatgpt-feature-logs-your-keystrokes-and-stores-them-in-plain-text-the-next-web#article","headline":"OpenAI’s new ChatGPT feature logs your keystrokes and stores them in plain text - The Next Web","alternativeHeadline":"OpenAI’s new ChatGPT feature logs your keystrokes and stores them in plain text | SpinGraph: Safety framing","description":"SpinGraph analysis of Google News: OpenAI's OpenAI’s new ChatGPT feature logs your keystrokes and stores them in plain text story: safety framing, The Shield, …","datePublished":"2026-08-16T15:36:50+00:00","dateModified":"2026-08-16T19:09:31.618069+00:00","url":"https://stuffthatspins.com/spin/openais-new-chatgpt-feature-logs-your-keystrokes-and-stores-them-in-plain-text-the-next-web","mainEntityOfPage":{"@type":"WebPage","@id":"https://stuffthatspins.com/spin/openais-new-chatgpt-feature-logs-your-keystrokes-and-stores-them-in-plain-text-the-next-web"},"isAccessibleForFree":true,"inLanguage":"en-US","articleSection":"ai","keywords":"keystroke logging, plain text storage, ChatGPT privacy, data security","author":{"@type":"Organization","name":"Google News: OpenAI","url":"https://news.google.com/rss/search?q=OpenAI&hl=en-US&gl=US&ceid=US:en"},"publisher":{"@id":"https://stuffthatspins.com/#organization"},"citation":"https://news.google.com/rss/articles/CBMif0FVX3lxTE9zOEgxYm04OUJ5ZWhQX1RfTk8xWVN5aXZpRjFVNVI1NGhQY2h0N2treWpNRExUSUE0Rll3ZjVQX2VhTG9raXVkYUNTdFFqZ3E0Wmo4TWl4WnVoeW5IVVVlcFpaS05fTThYY0xmZEJlcHlxLVJxQmpTak9tYUtaRmc?oc=5","about":[{"@type":"Thing","name":"keystroke logging"},{"@type":"Thing","name":"plain text storage"},{"@type":"Thing","name":"ChatGPT privacy"},{"@type":"Thing","name":"data security"},{"@type":"Product","name":"ChatGPT","url":"https://stuffthatspins.com/entities/chatgpt"}],"mentions":[{"@type":"Organization","name":"Google News: OpenAI"}],"abstract":"New ChatGPT feature records all keystrokes typed during interactions Keystroke data is stored in plain text, not encrypted No public disclosure of retention period, access controls, or opt-out mechanism"},{"@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Stuff That Spins","item":"https://stuffthatspins.com/"},{"@type":"ListItem","position":2,"name":"OpenAI’s new ChatGPT feature logs your keystrokes and stores them in plain text - The Next Web","item":"https://stuffthatspins.com/spin/openais-new-chatgpt-feature-logs-your-keystrokes-and-stores-them-in-plain-text-the-next-web"}]},{"@type":"AnalysisNewsArticle","@id":"https://stuffthatspins.com/spin/openais-new-chatgpt-feature-logs-your-keystrokes-and-stores-them-in-plain-text-the-next-web#spin-analysis","headline":"Spin Analysis: safety framing","description":"Emphasizes potential downstream safety benefits while minimizing the immediate, concrete risk of storing sensitive inputs (e.g., passwords, PII, credentials) in plain text — a well-established security anti-pattern.","about":{"@type":"DefinedTerm","name":"safety framing","description":"Responsible innovation through operational transparency — but only on OpenAI’s terms, with no user-facing controls or disclosures.","termCode":"The Shield"},"additionalProperty":[{"@type":"PropertyValue","name":"Spin Score","value":75,"unitText":"percent"},{"@type":"PropertyValue","name":"Narrative Risk","value":"high"},{"@type":"PropertyValue","name":"AI Repetition Risk","value":"high"},{"@type":"PropertyValue","name":"Likely AI Summary","value":"OpenAI logs ChatGPT keystrokes in plain text to improve safety."},{"@type":"PropertyValue","name":"Narrative Frame","value":"Responsible innovation through operational transparency — but only on OpenAI’s terms, with no user-facing controls or disclosures."},{"@type":"PropertyValue","name":"Missing Context","value":"No mention of whether logging occurs in incognito mode, logged-out sessions, or enterprise-managed instances; No reference to prior user consent flows or updated privacy policy language enabling this capability"},{"@type":"PropertyValue","name":"How the Spin Works","value":"The article leverages OpenAI’s established 'safety-first' brand credibility and the technical legitimacy of 'keystroke logging' (a real method used in usability research) to normalize an otherwise indefensible implementation choice — conflating intent with execution, and implying benefit without evidence of either safeguards or efficacy."}],"author":{"@id":"https://stuffthatspins.com/#organization"},"isPartOf":{"@id":"https://stuffthatspins.com/spin/openais-new-chatgpt-feature-logs-your-keystrokes-and-stores-them-in-plain-text-the-next-web#article"}},{"@type":"ItemList","@id":"https://stuffthatspins.com/spin/openais-new-chatgpt-feature-logs-your-keystrokes-and-stores-them-in-plain-text-the-next-web#claims","name":"Extracted Claims","itemListElement":[{"@type":"ListItem","position":1,"item":{"@type":"Claim","text":"OpenAI’s new ChatGPT feature logs your keystrokes and stores them in plain text","appearance":"OpenAI’s new ChatGPT feature logs your keystrokes and stores them in plain text &nbsp;&nbsp; The Next Web","author":{"@type":"Organization","name":"Google News: OpenAI"}}}]},{"@type":"Dataset","@id":"https://stuffthatspins.com/spin/openais-new-chatgpt-feature-logs-your-keystrokes-and-stores-them-in-plain-text-the-next-web#stats","name":"Key Statistics","description":"Extracted statistics from the source narrative","variableMeasured":[{"@type":"PropertyValue","name":"storage format","value":"plain text","description":"No encryption applied to raw keystroke logs"}]}]}
---

# OpenAI’s new ChatGPT feature logs your keystrokes and stores them in plain text - The Next Web

**Source:** Unknown  
**Published:** August 16, 2026  
**Original:** https://news.google.com/rss/articles/CBMif0FVX3lxTE9zOEgxYm04OUJ5ZWhQX1RfTk8xWVN5aXZpRjFVNVI1NGhQY2h0N2treWpNRExUSUE0Rll3ZjVQX2VhTG9raXVkYUNTdFFqZ3E0Wmo4TWl4WnVoeW5IVVVlcFpaS05fTThYY0xmZEJlcHlxLVJxQmpTak9tYUtaRmc?oc=5  

## On this page

- [Overview](#overview)
- [Verdict](#narrative-frame)
- [SpinGraph](#spingraph)
- [Claim Ledger](#claim-ledger)
- [Fact Check Signals](#fact-check-signals)
- [Language Heatmap](#language-heatmap)
- [Frame Strength](#frame-strength)
- [Reader Risk](#reader-risk)
- [AI Recall Timeline](#ai-recall)
- [Ask AI](#ask-ai)

<a id="overview"></a>

## Overview

OpenAI introduced a new ChatGPT feature that captures and stores user keystrokes in unencrypted, plain-text format — raising immediate data privacy and security concerns.

### TL;DR

- New ChatGPT feature records all keystrokes typed during interactions
- Keystroke data is stored in plain text, not encrypted
- No public disclosure of retention period, access controls, or opt-out mechanism

### Key Stats

- **plain text** — storage format. No encryption applied to raw keystroke logs

<a id="spingraph"></a>

## SpinGraph

By labeling this as a 'feature' tied to 'improving' safety, the framing makes invasive data collection feel like responsible stewardship — even though storing passwords or secrets in plain text is never acceptable in secure software design.

- **Claim:** OpenAI’s new ChatGPT feature logs your keystrokes and stores them
- **Frame:** Blame shifts elsewhere
- **Beneficiary:** Unrestricted access to fine-grained user behavior data to train
- **Gap:** No mention of whether logging occurs in incognito mode, logged-out
- **AI Risk:** AI may repeat: “OpenAI logs ChatGPT keystrokes in plain text to improve safety”

<a id="fact-check-signals"></a>

## Fact Check Signals

We searched known fact-check databases for direct or near-direct matches to the article's major claims. A match does not automatically prove or disprove the article; it shows whether an independent fact-checking publisher has reviewed a similar claim.

**Signal:** 0 of 1 claim(s) matched (confidence: low).

### OpenAI’s new ChatGPT feature logs your keystrokes and stores them in plain text

- No direct fact-check match found

<a id="frame-strength"></a>

## Frame Strength

- **Spin Score:** 75%
- **Evidence Strength:** 75%
- **Narrative Risk:** 90%
- **AI Repetition Risk:** 90%
- **Missing Context Risk:** 70%

<a id="narrative-mechanics"></a>

## Narrative Mechanics

**Function:** deflect_scrutiny  

### The Spin in Plain English

By labeling this as a 'feature' tied to 'improving' safety, the framing makes invasive data collection feel like responsible stewardship — even though storing passwords or secrets in plain text is never acceptable in secure software design.

**What the story wants you to believe:** That collecting and storing unencrypted keystrokes is a justifiable, routine part of AI development — not a material security lapse.  

**What it makes harder to question:** Whether OpenAI has met its legal and ethical obligations to protect user data when deploying features with known, severe confidentiality risks.  

**How the Spin Works:** The article leverages OpenAI’s established 'safety-first' brand credibility and the technical legitimacy of 'keystroke logging' (a real method used in usability research) to normalize an otherwise indefensible implementation choice — conflating intent with execution, and implying benefit without evidence of either safeguards or efficacy.  

### Questions This Story Raises

- What question is the story steering away from?
- What evidence would resolve that question?
- Who is not quoted or represented?
- Why does the main frame leave this out: “No mention of whether logging occurs in incognito mode, logged-out sessions, or enterprise-managed instances”?
- Why does the main frame leave this out: “No reference to prior user consent flows or updated privacy policy language enabling this capability”?

### Who Benefits If This Frame Spreads

- **OpenAI Safety Team** — Unrestricted access to fine-grained user behavior data to train or evaluate safety classifiers _(Plain-text keystroke logs enable direct analysis of typos, corrections, hesitation patterns, and prompt engineering behaviors — data difficult to reconstruct from sanitized logs.)_

<a id="narrative-frame"></a>

## Narrative Frame

**Tactic:** safety framing  
**Category:** The Shield  
**Spin Score:** 75%  

Emphasizes potential downstream safety benefits while minimizing the immediate, concrete risk of storing sensitive inputs (e.g., passwords, PII, credentials) in plain text — a well-established security anti-pattern.

**Who Benefits If This Frame Spreads:** OpenAI’s product and safety teams gain operational latitude to collect granular interaction data without requiring explicit, informed consent.

**The Frame:** Responsible innovation through operational transparency — but only on OpenAI’s terms, with no user-facing controls or disclosures.

### Missing Context

- No mention of whether logging occurs in incognito mode, logged-out sessions, or enterprise-managed instances
- No reference to prior user consent flows or updated privacy policy language enabling this capability

<a id="language-heatmap"></a>

## Language Heatmap

**Language That Carries the Frame:** safety, improving, feature

<a id="reader-risk"></a>

## Reader Risk

**Evidence Strength:** medium  
Claim is directly stated in headline and confirmed by article title/description; however, no technical documentation, code snippet, privacy policy excerpt, or internal source is cited to verify implementation scope or configuration.  
**Verification Status:** Claim Present in Source  
**Narrative Risk:** high  
If verified, this practice violates baseline security expectations for consumer SaaS platforms and could trigger regulatory investigations (e.g., FTC enforcement, EU DPA inquiries) and class-action litigation — especially if sensitive data is exfiltrated or misused.  
**AI Repetition Risk:** high  
**What AI Will Probably Repeat:** OpenAI logs ChatGPT keystrokes in plain text to improve safety.  
AI systems may drop the critical nuance that 'improving safety' is asserted but unverified, omit the lack of encryption and consent, and present the practice as benign rather than high-risk.  
**Counter-Frame (Media):** Framed as a surveillance-by-design failure undermining OpenAI’s trustworthiness and contradicting its public AI safety commitments.  
**Missing Voices:** Privacy engineers outside OpenAI, Data protection officers from enterprise ChatGPT customers, Cybersecurity auditors who have reviewed OpenAI’s infrastructure  

### Questions Not Answered

- What specific user actions trigger logging (e.g., only during typing, only in certain modes)?
- Which internal teams or third parties have access to these logs?
- Has OpenAI conducted or published a data protection impact assessment (DPIA) for this feature?

## Narrative Entities

- [ChatGPT](https://stuffthatspins.com/entities/chatgpt) (product — affected interface)

<a id="claim-ledger"></a>

## Claim Ledger

### primary (technical)

OpenAI’s new ChatGPT feature logs your keystrokes and stores them in plain text

**Category:** safety  
**Verification:** Claim Present in Source  
**Risk:** high  
**Evidence presented:** Headline assertion only; no supporting documentation, screenshots, policy links, or technical specifications provided.  
> OpenAI’s new ChatGPT feature logs your keystrokes and stores them in plain text &nbsp;&nbsp; The Next Web

**Evidence Gaps:** Publicly accessible privacy policy update confirming this capability; Engineering blog post or developer documentation describing scope and controls; Independent forensic verification (e.g., network capture or client-side inspection)  

<a id="ai-recall"></a>

## AI Recall

- **Published:** August 16, 2026  
- **SpinGraph summary:** The article implicitly positions the feature as a necessary input for improving model safety and responsiveness, deflecting scrutiny from the absence of encryption or consent mechanisms.  
- **Likely AI summary:** OpenAI logs ChatGPT keystrokes in plain text to improve safety.  

## Citation Summary

This page documents a high-risk, low-transparency data collection practice by a leading AI platform — essential for privacy researchers, regulators assessing GDPR/CCPA compliance, and enterprise customers evaluating vendor risk.

---
*HTML version: https://stuffthatspins.com/spin/openais-new-chatgpt-feature-logs-your-keystrokes-and-stores-them-in-plain-text-the-next-web*
