AI Chatbots: Which Ones Collect the Most Data?

The rise of AI chatbots has revolutionized how people access information, create content, and communicate. However, this convenience often comes at the expense of user privacy. A recent report from Surfshark, a leading VPN and security service, sheds light on the extent of data collection by popular AI platforms, revealing which chatbots gather the most user information.

Surfshark analyzed the privacy details of 10 widely used AI chatbots, including ChatGPT, Claude AI, DeepSeek, Google Gemini, Grok, Jasper, Meta AI, Microsoft Copilot, Perplexity, Pi, and Poe. The study examined the types of data these platforms collect, whether the data is linked to users, and the involvement of third-party advertising.

Data Collection by AI Chatbots

The report evaluated 35 data types, ranging from basic contact information to sensitive details like health, financial, and location data. Surprisingly, all 10 AI chatbots collect some form of user data. On average, these apps gather 13 out of the 35 data types. Nearly 45% of the apps track location data, while almost 30% link user data with third-party sources for targeted advertising or data sharing.

Sensitive information, such as racial or ethnic data, sexual orientation, health records, and biometric data, is also collected by some AI platforms. This raises significant privacy concerns, as such data can be highly personal and vulnerable to misuse.

Worst Offenders in Data Collection

Among the AI chatbots analyzed, Meta AI stands out as the most aggressive data collector, amassing 32 out of 35 data types. This includes financial information, health and fitness data, and sensitive information. Meta AI is the only app that collects data across all major categories, making it a significant outlier in the study.

Google Gemini follows, collecting 22 different data types, including precise location, contact information, user content, and browsing history. Poe, Claude, and Copilot also gather substantial amounts of data, with Poe and Copilot using it to track users and display targeted ads.

DeepSeek collects 11 types of data, including chat history, which is sent to servers in China. This practice has led to a breach where over 1 million records were leaked, highlighting the risks associated with data storage and transmission.

ChatGPT, while gathering 10 types of data, offers privacy-focused features such as temporary chats that delete data after 30 days. Users can also request that their personal data not be used for training purposes, providing a level of control over their information.

Protecting Your Data

As AI chatbots become more integrated into daily life, it is crucial for users to understand how their data is being collected and used. While data collection is common among AI platforms, services like ChatGPT provide options to limit or prevent the gathering of personal information.

Experts recommend reviewing the privacy policies and settings of any AI service you use. By taking proactive steps, users can better protect their data and ensure their privacy is respected in an increasingly digital world.