AI incident ·
Reported Public Exposure of Over 100,000 LLM Conversations via Share Links Indexed by Search Engines and Archived
On this page 10 sections
In brief
An AI system built by xAI, OpenAI and 6 others and deployed by xAI, OpenAI and 4 others allegedly harmed Qwen users, Privacy and 7 others.
- Risk domain
- Privacy & Security
- Occurred
- Coverage
- 5 reports
What happened
Across 2024 and 2025, the share features in multiple LLM platforms, including ChatGPT, Claude, Copilot, Qwen, Mistral, and Grok, allegedly exposed user conversations marked "discoverable" to search engines and archiving services. Over 100,000 chats were reportedly indexed and later scraped, purportedly revealing API keys, access tokens, personal identifiers, and sensitive business data.
Editor's notes
Timeline note: According to reporting, a lot of the "discoverable" chats were archived by the Internet Archive by at least June 2024, but actual reporting on the incident did not appear to arise until the end of July and early August 2025. This incident ID was created 08/25/2025 to capture the range of reporting on this cluster of related harms. See also Incident 1172: Meta AI Bug in Deployed Service Reportedly Allowed Potential Access to Other Users' Prompts and Responses. See also: Incident 1621: Publicly Shared Claude Artifacts Reportedly Became Discoverable Through Google Search (https://incidentdatabase.ai/cite/1621).
Laws on record for this risk area
Policy angle: Classified under Privacy & Security (Compromise of privacy by obtaining, leaking or correctly inferring sensitive information) in the MIT AI Risk Repository taxonomy; 4 instruments on record cover this risk area.
- India DPDP Act
- Law No. 132/2025 on artificial intelligence
- EU AI Act
- Texas Responsible AI Governance Act (TRAIGA)
Matched from the record's risk category and country to the instruments in effect recorded here: a law listed covers the area, which is not a finding that it applies to this incident. A reviewer can correct the match in the repository (data/external/incident_overrides.yaml).
News reports (5)
Titles link to the original publisher; report text is not reproduced here.
Who was involved
- Alleged developer
- xAI OpenAI Mistral Microsoft Large language model developers Chatbot developers Anthropic Alibaba
- Alleged harmed party
- Qwen users Privacy Mistral users Information integrity Grok users General public Copilot users Claude users ChatGPT users
AI systems implicated
Wayback MachineQwenMistralLarge language modelsInternet ArchiveGrokGoogleCopilotClaudeChatGPTChatbots
Classification (MIT AI Risk Repository taxonomy)
- Risk domain
- Privacy & Security
- Risk subdomain
- 2.1 Compromise of privacy by obtaining, leaking or correctly inferring sensitive information
- Causal entity
- AI
- Intent
- Unintentional
- Timing
- Post-deployment
- Harm level
- —
- Sectors
- —
- Countries
- —
Risk entries describing this failure mode
Entries from the MIT AI Risk Repository coded to subdomain 2.1.
- Risks to privacy
"General- purpose AI models or systems can ‘leak’ information about individuals whose data was used in training. For future models trained on sensitive personal data like health or financial data, this may lead to partic...
- Risks to privacy
"General- purpose AI systems can cause or contribute to violations of user privacy. Violations can occur inadvertently during the training or usage of AI systems, for example through unauthorised processing of personal d...
- Privacy Leakage
"The model is trained with personal data in the corpus and unintentionally exposing them during the conversation."
- Association in LLMs
"Association in LLMs refers to the capability to associate various pieces of information related to a person. According to [68], [86], given a pair of PII entities (xi , xj ), which is associated by a model F. Using a pr...
- Privacy Leakage
"Privacy Leakage means the generated content includes sensitive personal information"
- Private Training Data
"As recent LLMs continue to incorporate licensed, created, and publicly available data sources in their corpora, the potential to mix private data in the training corpora is significantly increased. The misused private d...
- Memorization in LLMs
"Memorization in LLMs refers to the capability to recover the training data with contextual prefixes. According to [88]–[90], given a PII entity x, which is memorized by a model F. Using a prompt p could force the model...
- Privacy and regulation violations
"Some of the broken systems discussed above are also very invasive of people’s privacy, controlling, for instance, the length of someone’s last romantic relationship [51]. More recently, ChatGPT was banned in Italy over...
Related incidents
Linked by editors or by text similarity in the source dataset.
Incidents in the same risk subdomain
- Grok Reportedly Disclosed Adult Performer Siri Dahl's Legal Name and Birthdate, Allegedly Contributing to Doxxing and Harassment
- NPR Host David Greene Alleged Google's NotebookLM Replicated His Voice Without Consent, Prompting Lawsuit
- Border Patrol Agent Allegedly Claimed Facial Recognition Identified Minneapolis ICE Observer and Global Entry Was Reportedly Revoked Three Days Later
- Perplexity AI Reportedly Accused in Federal Lawsuit of Purported Copyright Infringement and False Attribution of Chicago Tribune Content
- Secret Desires AI Platform Reportedly Exposed Nearly Two Million Sensitive Images in Cloud Storage Leak
- ChatGPT Reportedly Found to Reproduce Protected German Lyrics in Copyright Case
Other incidents involving xAI
- Grok Reportedly Generated Images After X Users Asked It to Unredact Epstein-File Photos of Children and Women
- X Users Reportedly Prompted Grok to Sexualize Images of Renée Good After Her Killing in Minneapolis
- xAI Reportedly Operated Unpermitted Gas Turbines in Southaven, Mississippi, to Power Colossus 2 AI Data Center, Producing Air Pollution and Noise
- xAI Allegedly Operates Unpermitted Methane Turbines in Memphis to Power Supercomputer Colossus to Train Grok
Source record: incident #1186 on the AI Incident Database · all 5 reports