AI Data Scraping Audit: What 6 Major Apps Collected From You in 2025
5 min readBy PrivateAI Team
AI Data Scraping Audit: What 6 Major Apps Collected From You in 2025
CONCLUSIONS:
By 2025, six major apps—Gmail, Dropbox, LinkedIn, Slack, Notion, and GitHub Copilot—will have collected vast amounts of personal and professional data, often without explicit user consent. This data is increasingly used to train AI systems, raising significant privacy and security concerns. The urgency for zero-knowledge alternatives is clear, as users demand greater control over their data. For actionable steps to prevent your content from being ingested into foundation models across these ecosystems, see our complete guide to opting out of AI training data.
1. Gmail:
- Data Collected: Subject lines, email content, attachments, and metadata (e.g., timestamps, device IDs).
- AI Use: Gmail's AI (like Google Assistant) uses this data to suggest responses, prioritize emails, and target ads.
- Risk: Users may unknowingly share sensitive information (e.g., financial or health-related emails) with Google's AI systems.
- Impact: Privacy breaches could expose personal or business communications. To understand how email ingestion fuels commercial models and how to migrate away, see our analysis on why your email is an AI training dataset.
2. Dropbox:
- Data Collected: File names, content (if stored unencrypted), and access patterns.
- AI Use: Dropbox integrates with AI tools (e.g., Copilot) to enhance file organization and collaboration.
- Risk: Encrypted files may still be analyzed for metadata, potentially revealing sensitive information.
- Impact: Misuse of file data could lead to unauthorized access or leaks. Teams handling confidential records should replace unencrypted cloud storage with zero-knowledge architectures detailed in our Tresorit vs. Proton Drive encrypted storage comparison.
3. LinkedIn:
- Data Collected: User profiles, connections, job history, and engagement metrics.
- AI Use: LinkedIn's AI algorithms analyze this data to personalize job recommendations and ads.
- Risk: Sensitive career-related information could be used for targeted recruitment or hiring bias.
- Impact: Discrimination or unfair opportunities based on AI-driven insights. To restrict scraping and control who accesses your employment history, follow our LinkedIn privacy lockdown guide.
4. Slack:
- Data Collected: Chat history, file attachments, and user activity logs.
- AI Use: Slack's AI features (e.g., Copilot) analyze conversations to suggest responses and identify trends.
- Risk: Sensitive business communications could be exposed or misused.
- Impact: Breaches of confidentiality in workplace discussions. To safeguard sensitive attachments and internal records, implement our protocol to stop sending sensitive files in Slack.
5. Notion:
- Data Collected: Notes, task lists, and database entries, often synced across devices.
- AI Use: Notion's AI tools analyze content to generate templates, organize tasks, and provide insights.
- Risk: Sensitive personal or professional notes could be accessed by third-party AI systems.
- Impact: Loss of control over structured data, leading to potential misuse.
6. GitHub Copilot:
- Data Collected: Code snippets, commit history, and coding patterns.
- AI Use: Copilot uses this data to generate code suggestions and improve developer productivity.
- Risk: Proprietary code could be shared or reverse-engineered by AI systems.
- Impact:INFRINGEMENT or unauthorized use of intellectual property.