AI Stealing Data: Nowadays, we are all connected to AI technologies in some form or the other, be it AI assistants like ChatGPT or Microsoft Copilot or our fitness tracking through smartwatches.
AI Stealing Data: Nowadays, we are all connected to AI technologies in some form or the other, whether it is AI assistants like ChatGPT or Microsoft Copilot or our fitness tracking through smartwatches. While these technologies make our lives easier, they also raise serious questions about our privacy. Christopher Ramzan, Assistant Professor of Cybersecurity at West Virginia University, has done research on how modern AI systems collect our personal information and how we can protect our privacy from them.
How do AI tools collect your information?
Christopher explains that in generative AI like ChatGPT and Google Gemini, everything you type—questions, answers, or suggestions—is recorded, stored, and used to improve the model. Although platforms like OpenAI give you the option to not use the data for training, the information you input is still stored. Many companies claim to anonymize your data, but there is always a risk that the data can be re-identified.
Predictive AI and Social Media
In addition to generative AI, social media platforms such as Facebook, Instagram, and TikTok also constantly analyze your behavior. Every post, photo, video, like, share, comment, and even how long you looked at something helps AI systems create a digital profile of you. Smart devices such as smartwatches, fitness trackers, and home speakers also constantly collect data from biometric data, voice recognition, and location tracking.
A threat to your privacy
Ramzan explains that the data of AI tools is often stored in the cloud. This means that a third party can also access that data. The information recorded from smartwatches or voice devices is used to improve AI algorithms or create user profiles. This has a direct impact on data privacy laws and user security.
How secure is your information?
The biggest problem is the lack of transparency. People do not know what data is being collected, how it is being used and with whom it is being shared. Companies create privacy policies written in complex language which is difficult for common users to read and understand. According to a report, on an average people read the Terms of Service and agree to it in just 73 seconds whereas it can take more than 30 minutes to understand it properly.
Even though AI companies are trustworthy, their data centers are not completely safe from the threat of cyber attacks. Hackers or foreign agencies can target these devices and servers and steal your sensitive information.
What should you do to protect your privacy?
Ramzan gives some important tips
- Never enter personal information like name, date of birth, address, Aadhaar/PAN number etc. in AI tools.
- When using AI in the office, do not share confidential information, such as trade secrets or client data.
- Never tell AI tools something that you wouldn’t want to make public.
- Turn off smart devices when they are not needed. Simply being in "sleep mode" is not enough. Devices are often listening in the background.
- Read the privacy policy carefully. Know what you have already allowed.
