Summary
**Surfshark** has ranked 10 AI chatbots based on data collection practices, revealing **Meta AI** collects 18 data types while **ChatGPT** logs 12. The analysis [[~ai-privacy|AI privacy]] focuses on Apple App Store disclosures and privacy policies, highlighting **DeepSeek**'s 14 data points. [[~data-collection|Data collection]] practices vary wildly, with some apps retaining data for 18 months. [[~chatgpt|ChatGPT]]'s 2025 updates show increased transparency, but **Google Gemini**'s 16 data types raise concerns. This ranking [[~surfshark|Surfshark]] underscores the need for user awareness in an era of pervasive AI adoption. [[~digital-democracy|Digital democracy]] hinges on understanding these invisible data flows. [[~privacy-policies|Privacy policies]] remain opaque for many apps, creating a regulatory gray zone.
Key Takeaways
- Surfshark's analysis reveals stark differences in AI chatbot data collection practices
- Meta AI collects the most data (18 types) while ChatGPT shows improved transparency
- Data retention periods vary significantly between platforms
- Apple App Store disclosures provide critical but incomplete information
- Standardized AI privacy frameworks are urgently needed
Balanced Perspective
**Surfshark** analyzed 10 AI chatbots based on Apple App Store disclosures and privacy policies, focusing on data types collected and retention periods. **Meta AI** logs the most data (18 types), while **ChatGPT** collects 12. [[~deepseek|DeepSeek]] falls in between with 14. The study notes **Google Gemini**'s 16 data points and **ChatGPT**'s 2025 updates. [[~privacy-policies|Privacy policies]] remain inconsistent, with some apps retaining data for 18 months. This analysis provides a factual baseline for user decision-making without overt advocacy.
Optimistic View
**Surfshark's** ranking empowers users to choose AI tools with clearer privacy frameworks. **ChatGPT**'s 2025 updates show progress toward transparency, while **DeepSeek**'s 14 data points reflect a balanced approach. [[~ai-ethics|AI ethics]] frameworks could now be tailored to these specific data collection models. The **Apple App Store**'s disclosure requirements are a step forward, but more needs to be done to standardize [[~data-retention|data retention]] policies. This analysis could spur industry-wide reforms, creating safer digital ecosystems for users.
Critical View
**Surfshark's** ranking exposes a dangerous lack of standardization in AI data practices. **Meta AI**'s 18 data points include location and device information, creating privacy risks. [[~data-retention|Data retention]] periods of 18 months for some apps are alarming. **ChatGPT**'s 2025 updates show progress, but **Google Gemini**'s 16 data types still raise red flags. The **Apple App Store**'s disclosures are insufficient, leaving users vulnerable to data exploitation. This gap in [[~ai-privacy|AI privacy]] could enable corporate surveillance on an unprecedented scale.
Source
Originally reported by surfshark.com