Research by Maple · CC BY 4.0
AI Privacy Timeline
Every chat you send an AI company lives under a policy that can change, on a server that can leak, inside a company a court can order to keep it. This is the record of each time that happened.
Compiled and maintained by
72 entries since March 24, 2023, each tied to one primary source. Data through September 21, 2026, checked October 16, 2026. Original at maple.ai/privacy-timeline.
40
Policy changes
10
Breaches and leaks
10
Overreach
12
Regulation
01 / The record
What happened to private data.
Newest first. Each entry links to its primary source and to the same entry on Maple’s timeline. For the ChatGPT side of this in plain terms, read our guide on whether ChatGPT is private.
202623 entries
Breaches and leaks
Bee Cheng Hiang
AI-written email script exposed 95,364 addresses
Singapore's data regulator said an employee's generative AI email script exposed 95,364 members' email addresses to other recipients in an April 25, 2026 campaign.
Policy changes
Anthropic
Korea addendum shares identity-verification data
For users covered by Korea's privacy law, identity-verification data may be shared with the Korea Credit Bureau.
Policy changes
OpenAI
Past chats can personalize ChatGPT ads
When ad personalization is available and on, past chats are used to make ads more relevant for Free and Go users.
Policy changes
xAI
Grok policy names SpaceXAI as controller
The Grok privacy policy names SpaceXAI LLC as the controller for the Grok app and grok.com.
Policy changes
Microsoft
Copilot ads may use chat history
If ads personalization is on, personalized ads may use chat history, saved memories, and other Microsoft data you chose to share.
Your control Settings > Personalization > Allow ads personalization.
Policy changes
Microsoft
Copilot app narrows its training statement
New wording says prompts, responses and file contents aren’t used to train “foundation models”; the older FAQ said “generative AI models.”
Policy changes
Atlassian
Atlassian starts AI training on cloud data
Cloud customer metadata and in-app content, such as Jira and Confluence text, began to be used to improve Atlassian's AI; metadata opt-out is Enterprise-only.
Overreach
European Union
EU restores voluntary message scanning law
Regulation 2026/1881 was adopted, again allowing messaging providers to scan communications for child sexual abuse until April 2028, excluding end-to-end encrypted communications.
Regulation
Character.AI
Italy fines Character.AI €158,000
Italy's Garante fined Character Technologies €158,000 over its privacy notice, a late impact assessment, and age verification for minors.
Policy changes
Anthropic
Claude policy adds age checks and sharing
Consumer accounts may be asked for age or identity data, and multi-step tasks can share inputs and outputs with connected apps.
Policy changes
Microsoft
Copilot in Edge can read page content
Copilot in Edge may access page content, tabs, browsing history, and visual snapshots of the page.
Policy changes
Microsoft
Copilot Health uses health information
Copilot Health uses health information to provide personalized assistance focused on health and wellness.
Regulation
European Union
EU agrees to delay AI Act high-risk rules
Lawmakers agreed to move AI Act high-risk deadlines to December 2027 and August 2028 and to ban AI nudifier apps.
Policy changes
xAI
Google app data excluded from training
Content from Google apps connected through Google OAuth is not used for internal AI or model training.
Policy changes
xAI
Grok policy adds targeted advertising
The Grok privacy policy adds delivering targeted advertising as a listed use of personal information.
Policy changes
OpenAI
ChatGPT adds controls for personalized ads
Account settings include controls over what data is used to personalize ads shown on the services.
Your control Advertising controls in your account settings.
Policy changes
OpenAI
Parents can link ChatGPT teen accounts
Teen users and a parent or guardian can link accounts so the adult can manage settings and receive safety alerts.
Policy changes
GitHub
GitHub to train on Copilot interaction data
Blog post: Copilot Free, Pro and Pro+ interaction data, including code snippets, would train GitHub's AI models from April 24 unless users opt out.
Your control Opt out in Copilot settings under "Privacy."
Policy changes
Microsoft
Statement drops stated retention periods
The privacy statement no longer describes specific retention periods in its Bing and Advertising sections.
Overreach
Anthropic
Court rules Claude chats not privileged
A federal judge ruled that a criminal defendant's written exchanges with Anthropic's Claude were not protected from government inspection by attorney-client privilege or work product.
Regulation
X and xAI
UK ICO opens investigation into Grok
The UK ICO opened formal investigations into X and xAI over Grok's processing of personal data and its generation of sexualized images.
Policy changes
Google
Gemini app connections start off by default
Connecting Gmail, Photos, YouTube, and Search is off by default, and Gemini does not train directly on Gmail or Photos.
Overreach
OpenAI
Judge upholds 20 million ChatGPT log order
A federal district judge denied OpenAI's objections and affirmed orders requiring production of 20 million de-identified ChatGPT conversation logs.
202523 entries
Overreach
Council of the EU
EU Council agrees position on Chat Control
Member states agreed the Council position on the child sexual abuse regulation, including making permanent an exemption that lets companies voluntarily scan messages.
Breaches and leaks
OpenAI
Mixpanel breach exposed OpenAI API user details
An attacker exported data from analytics vendor Mixpanel, including names, emails, and approximate locations of some OpenAI API users; chats were not exposed.
Overreach
OpenAI
Court orders 20 million ChatGPT logs produced
A federal magistrate judge ordered OpenAI to produce 20 million de-identified consumer ChatGPT logs to news plaintiffs in the copyright litigation.
Regulation
State of California
California enacts companion chatbot law SB 243
The governor signed SB 243, requiring companion chatbots to disclose they are AI and maintain self-harm response protocols, with protections for minors.
Overreach
OpenAI
Court ends OpenAI's ongoing chat preservation duty
A court-approved stipulation ended OpenAI’s ongoing duty to preserve deleted ChatGPT output logs as of September 26, 2025; earlier logs stay, except EEA, Swiss and UK data, and listed-domain accounts remain preserved.
Policy changes
Meta
Meta to use AI chats for ads
Newsroom post: content and ad recommendations would be personalized based on people's interactions with Meta's generative AI features, starting December 16, 2025.
Overreach
OpenAI
Federal warrant issued for a user’s ChatGPT records
A federal magistrate judge in Maine issued a warrant for OpenAI records of a user’s ChatGPT searches; the return shows it was executed the next day.
Regulation
Federal Trade Commission
FTC orders seven AI chatbot firms to report
6(b) orders went to seven companies, including OpenAI, Meta, and Character.AI, on how their companion chatbots affect children and teens.
Breaches and leaks
Salesloft
Drift chatbot OAuth tokens stolen and misused
Trust center update: an attacker stole OAuth tokens for Drift AI chat integrations and used them to take customer data, including from Salesforce in August 2025.
Policy changes
Anthropic
Claude consumer chats may train models
Claude Free, Pro and Max users were asked to choose whether chats train Anthropic's models; retention extends to five years for those who allow it.
Your control Change your choice at any time in Privacy Settings.
Overreach
OpenAI
OpenAI may refer some chats to police
Company post: human reviewers examine conversations flagged for planned harm to others, and OpenAI may refer imminent threats of serious physical harm to law enforcement.
Policy changes
Google
Gemini to sample user uploads for improvement
Blog post: Gemini Apps Activity would be renamed Keep Activity and, when on, a sample of future uploads used to improve Google services.
Your control Turn off Keep Activity, or use Temporary Chats.
Breaches and leaks
Change Healthcare
Change Healthcare breach reaches 192.7 million people
HHS's Office for Civil Rights was told that about 192.7 million people were affected by Change Healthcare's 2024 ransomware attack.
Policy changes
xAI
Grok training objection moves into settings
People in Europe can object in settings to use of their information to train Grok models.
Your control Object to model training in your settings (people in Europe).
Breaches and leaks
Paradox.ai
Hiring chatbot test account exposed applicant chats
Researchers used a legacy test account password to reach chat records in one client's AI hiring assistant, viewing five U.S. candidates' contact details.
Regulation
Luka (Replika)
Italy fines Replika maker €5 million
Italy's Garante fined Luka Inc. €5 million over Replika's legal basis for processing, its privacy policy, and lack of age verification.
Overreach
OpenAI
Court orders OpenAI to keep deleted chats
A federal magistrate judge ordered OpenAI to preserve and segregate all ChatGPT output log data that would otherwise be deleted, including at users' request.
Policy changes
Meta
Meta resumes AI training on EU content
Newsroom post: Meta's AI would be trained on public posts and comments from adults in the EU, and on their interactions with Meta AI.
Your control Objection form, linked from the notices EU users receive.
Overreach
Apple
Apple pulls Advanced Data Protection in the UK
Advanced Data Protection, Apple's end-to-end encryption for iCloud, stopped being offered to new users in the United Kingdom.
Policy changes
xAI
Parents must agree for teenage Grok use
Parents of teenagers ages 13 to 17 must agree to the terms before those teenagers use the service.
Regulation
European Union
EU AI Act prohibitions take effect
The AI Act's bans on prohibited practices, such as social scoring and workplace emotion recognition, began to apply.
Regulation
DeepSeek
Italy orders DeepSeek to stop processing data
Italy's Garante ordered DeepSeek's two Chinese operators to stop processing Italian users' data, effective immediately, and opened an investigation.
Policy changes
xAI
Temporary Grok chats kept for 30 days
Temporary chats do not appear in history and are kept for up to 30 days for safety purposes.
202415 entries
Regulation
OpenAI
Italy fines OpenAI €15 million
Italy's Garante fined OpenAI €15 million over ChatGPT's legal basis for training, transparency, and breach notification; a Rome court later upheld OpenAI's appeal.
Policy changes
xAI
Grok minimum age lowered to 13
The minimum age changed from 18 to 13, and the service is not directed at children under 13.
Policy changes
X
X lets third parties train AI on data
Privacy policy update: third parties receiving X data may use it for their own purposes, including training their AI models.
Policy changes
xAI
Grok interaction data added to training
Interaction data, such as session frequency and features used, is listed among data used to train models.
Policy changes
OpenAI
Temporary chats kept up to 30 days
Temporary chats do not appear in history and are kept for up to 30 days for safety purposes.
Policy changes
LinkedIn
LinkedIn uses member data for generative AI
Privacy policy update: member data trains LinkedIn's content-generating AI models, with an opt-out setting; the EEA, Switzerland and UK were excluded.
Your control Opt-out setting for generative AI training.
Policy changes
xAI
User content may train Grok models
User content provided to the service may be processed to train and improve the models, with an email opt-out.
Regulation
European Union
EU AI Act enters into force
The Artificial Intelligence Act entered into force, setting risk-based rules for AI systems that apply in stages over the following years.
Policy changes
Microsoft
Privacy statement says data may train models
The privacy statement says Microsoft may use personal data to develop and train its AI models.
Policy changes
Meta
Meta plans AI training on Europeans' posts
Newsroom post: Meta's AI models would be trained on public posts, comments and photos from adult users in Europe, with an objection form.
Policy changes
Anthropic
Deleted Claude chats removed within 30 days
Deleted conversations are removed from history immediately and deleted from back-end systems within 30 days.
Breaches and leaks
Hugging Face
Hugging Face detects access to Spaces secrets
Unauthorized access to secrets stored in the Spaces AI app platform was detected; affected tokens were revoked, and users and data protection authorities notified.
Breaches and leaks
Kaiser Permanente
Kaiser tracking pixels sent member data to advertisers
Members were notified that tracking pixels on Kaiser's sites and apps may have sent names, IP addresses, and health searches to Google, Bing, and X.
Policy changes
xAI
Grok training covers public posts and chats
Training may use public X posts, internet search data, and a voluntary subset of conversation and feedback data.
Policy changes
Anthropic
Claude chats excluded from training by default
Inputs and outputs are not used to train models unless a chat is safety-flagged, reported, or the user opts in.
202311 entries
Policy changes
OpenAI
Business admins can access ChatGPT accounts
Enterprise or business-account administrators may access and control an account, and users can share conversations by link.
Policy changes
xAI
Early policy trains only on chosen data
Models are trained only on usage data people choose to provide, not on account, payment, or technical data.
Breaches and leaks
23andMe
23andMe accounts accessed via credential stuffing
Attackers used reused passwords to access about 14,000 accounts, reaching data from about 5.5 million DNA Relatives and 1.4 million Family Tree profiles.
Policy changes
X
X privacy policy adds AI model training
Privacy policy update: the information X collects and publicly available information may be used to help train its machine learning or AI models.
Breaches and leaks
Microsoft
AI research token exposed internal Microsoft data
An employee published a storage link with a broad-access SAS token on GitHub, exposing two former employees' workstation backups and internal Teams messages.
Policy changes
Microsoft
Bing Chat activity can be viewed and managed
The Bing section explains how to view and manage recent activity with Bing Chat.
Policy changes
Google
Google policy covers public data for AI training
Privacy policy update: wording changed from training "language models" on publicly available information to training "Google's AI models," naming Bard and Cloud AI.
Regulation
Amazon
FTC and DOJ charge Amazon over Alexa recordings
The FTC and DOJ alleged Amazon kept children's Alexa voice recordings indefinitely; a proposed order required a $25 million civil penalty and data deletion.
Policy changes
OpenAI
Turning off chat history stops model training
Conversations started with chat history off are not used to train models and are kept 30 days for abuse review.
Your control Turn off chat history in ChatGPT settings.
Regulation
OpenAI
Italy temporarily limits ChatGPT data processing
Italy's Garante ordered an immediate temporary limitation on OpenAI's processing of Italian users' data and opened an inquiry into ChatGPT.
Breaches and leaks
OpenAI
ChatGPT bug showed other users' chat titles
A Redis library bug let some ChatGPT users see other users' chat titles and exposed payment details of 1.2% of active Plus subscribers.
02 / The version you can check
Private by code, not by a policy that can change.
Maple cannot see your chats. Your messages are encrypted on your device and decrypted only inside a secure enclave your app checks before it sends a word.
Try Maple freeMaple co-founder Mark Suman walked through why AI chats became the next surveillance target and how private AI agents should work on the TFTC podcast.
03 / Credit
Built by Maple. Shared with credit.
This timeline is Maple’s research. Maple’s team finds each entry, ties it to one primary source, and has a person review it before it goes live. We republish it under the Creative Commons Attribution 4.0 license, which lets anyone share the work as long as Maple gets the credit.
- Title
- AI privacy timeline
- Created by
- Maple (Maple Privacy Labs)
- Original
- maple.ai/privacy-timeline
- Our changes
- TFTC changed the layout, added the category filter, and wrote the introduction, the questions and this note. Every entry’s headline, summary, control and source link is Maple’s, word for word.
- Data
- Through September 21, 2026, last checked by Maple on October 16, 2026. Download it from Maple as JSON or CSV, follow new entries by RSS, or read how entries are chosen.
- Corrections
- Maple keeps the record, so send suggestions and corrections to them through the original timeline. Fixes they make show up here within a day.
Citing it? Use Maple’s credit line
Source: Maple, AI privacy timeline (maple.ai/privacy-timeline), data through 2026-09-21.
04 / Questions
Common questions.
Who made this AI privacy timeline?
How are entries chosen?
Which AI company shows up most often?
Can I reuse this data?
Is there an AI app that can't read my chats?
Timeline entries are the work of Maple, republished under CC BY 4.0; primary sources keep their own terms. Not legal advice.
The Commoner
Truth for the Commoner, every weekday. Money, machines, and the people trying to control both.
Independent writing by Marty Bent at TFTC since 2017. Money, markets, AI, energy and privacy, delivered free to your inbox.
Free, every weekday. Unsubscribe anytime using the link in each newsletter. By subscribing you agree to our Terms and acknowledge our Privacy Policy. Read recent issues.