How to Protect Your Privacy While Using AI

The artificial intelligence (AI) revolution has spawned a wave of creativity and productivity tools, including chatbots, image generators and AI-powered search. These technologies are often neatly wrapped in a pristine package, missing little but a bow on top.

Having all this in one place has driven millions to platforms like ChatGPT and Gemini. People have started using these platforms to help with their finances, whether it's asking for estate planning advice, budget-setting suggestions or to talk through investment ideas.

But these tools are not without downsides. Users sometimes treat AI like friends or lovers, divulging their deepest, darkest secrets — including details of their financial situation. Yet, scrubbing one's personal information from these platforms is often difficult, if not impossible.

Subscribe to Kiplinger’s Personal Finance

Be a smarter, better informed investor.

Save up to 74%

Sign up for Kiplinger’s Free E-Newsletters

Profit and prosper with the best of expert advice on investing, taxes, retirement, personal finance and more - straight to your e-mail.

Profit and prosper with the best of expert advice - straight to your e-mail.

Sign up

While AI can be intriguing and even helpful, staying safe is crucial. Let's discuss this concern further and how to protect your privacy while using AI.

Does AI collect your data?

Every AI system is different, but most collect various data from you. ChatGPT, Gemini and the most popular AI platforms are data-hungry machines. They use large language models (LLMs) to answer our most burning questions. However, LLMs need massive amounts of data to create better models.

Due to this need for data, most AI tools store your chats by default and, in many cases, your location data and other information. The information AI collects might include:

• Your prompts or chats
• Information from your device, such as location data and phone numbers
• Your interests and preferences

John Licato, an associate professor at the University of South Florida's Bellini College of Artificial Intelligence, Cybersecurity and Computing, tells Kiplinger data collection is likely. "If an AI product is being hosted by a third-party (such as ChatGPT accessed through a web browser), then it's safe to assume that they are collecting data," Licato said.

Some AI tools let you toggle off the sharing of personal data, but that isn't always an option. When in doubt, it's best to err on the side of sharing less.

Is it safe to give ChatGPT or other AI personal information?

Sharing too much information with AI can be problematic. Sometimes, you may need to share details with AI to garner your desired response. However, there are no guarantees your information will be safe on these platforms.

"It's one thing if I asked AI to plan which restaurants to visit on my vacation to Thailand, but I'd never tell it my credit card information, or my name, or my address," said Joe Warnimont, senior analyst at HostingAdvice. Warnimont says this is due to a lack of knowledge about how the company behind the tool uses that information or who has access to it.

A Stanford Human-Centered Artificial Intelligence (HAI) report highlights a study that found training data could be recovered from ChatGPT and other models. In other words, someone could access your information without your consent or knowledge. While these tools are helpful, giving them too much personal information isn't safe.

Are Perplexity or other AI searches private?

Perplexity and other AI searches are not entirely private, as with most AI tools. Some may claim to value your privacy, but a quick look at their privacy policies reveals a familiar tune.

Perplexity's privacy policy states that it collects your input and output when using the service. This content “may constitute or contain personal information.” It may be stored, reproduced or published if you make it publicly available or share it with third parties.

Using AI search often creates the same concerns as other AI tools. "Perplexity and other AI-based search engines use large language models (LLMs) to do their heavy processing, just like other popular chat-based AI systems like ChatGPT," Licato says.

4 ways to stay safe when using AI

AI tools often collect a plethora of information about you by default. Follow these steps to stay safe while using these tools.

Check the company's privacy policy
The privacy policies of these tools describe what they store. Both policies have a section that says they collect chats and files you share, such as photos, screenshots or audio.

ChatGPT says it collects your location, such as data from your device's GPS. Google uses vaguer language, saying "info about your location includes the general area from your device." However, it also says it collects your IP address and work and home addresses stored in your Google account.

Both companies collect myriad details about you. Before using an AI tool, you must understand what data companies developing AI tools collect and be comfortable with the potential storage of that data.

Avoid sharing sensitive information
A fundamental principle of AI is to avoid sharing sensitive information when possible. While it may seem helpful or necessary to share specific details sometimes, less is often more. Details like your phone number, date of birth and address are information AI doesn’t need to know.

Sometimes, you may feed an AI basic information, such as your profession or how you feel that day. Giving AI these details isn't an enormous security risk, but remember that it may save your chats and files. If there is anything you don't want the public to know about you, keeping it private is best.

Kee Jefferys, co-founder of Session, a privacy-focused messaging app, recommends using placeholders. “If you frequently use AI tools, consider anonymizing sensitive information in your prompts,” Jefferys said. “This can be done by replacing personal or identifiable details with placeholder or generic information.”

So if you're asking AI for advice with a work problem, for example, Jefferys suggests using placeholders like “Company X” or “Founder X” instead of real names.

Adjust your settings
AI tools usually have settings you can adjust, though some give you more control than others. For instance, Gemini limits what you can control. It enables you to toggle whether you want to save certain information, such as your dietary preferences and the kind of language you want it to use. You can also control how long you want to save Gemini activity in Google Apps.

ChatGPT has more customization options, including privacy-related ones. For example, you can opt out of letting it use your data to train its models. You can also turn the memory feature off, meaning it won't "remember" details about you for future interactions. For extra security, you can enable multi-factor authentication (MFA) and quickly sign out on any additional devices you have.

Practice good security habits
Good security habits can keep you safe online, which applies to AI. For example, use strong passwords and MFA to keep your accounts safe. A password manager can help you use strong, secure passwords without remembering a long, complex stream of characters.

Remember that AI tools sometimes store information about us, including sensitive details. Keeping your account secure helps prevent unauthorized parties from accessing your information.

Learn more about AI

• How AI Will Impact Our Lives in 2025 and Beyond
• Major AI Companies You Should Know
• How AI Will Impact Your Workplace Retirement Plan
• What Is AI Investing?