How to prevent AI chatbots from collecting your data

“Hey ChatGPT, write me a thank you note for the guests who attended my party” or “Write me a summary paragraph based on my work experience” — getting work done is all fun and games until you realise that AI chatbots could actually be collecting information based on your queries.

After all, it isn’t exactly news that artificial intelligence is trained based on user data. But a few extra steps can help you prevent them from doing so. Here’s how to stop ChatGPT, Gemini or Copilot from collecting your data.

How to stop AI chatbots from accessing personal information

Before you do, it is crucial to note that AI-powered tools require user information and a vast amount of data to function efficiently. This is why, a lot of these companies make it especially difficult to opt out of data sharing.

However, this practice puts a spotlight on privacy and security and has raised multiple discussions about the use of personal content for development. So, if you’re someone who works with AI services regularly, you might want to check out these steps that ensure your information remains safe with you.

That said, take note that this will only stop the chatbots from accessing your data in the future, and will not work on anything shared previously.

ChatGPT

OpenAI’s ChatGPT recently witnessed a lot of online debates for introducing its ‘Memory’ feature. The feature not only allows the chatbot to remember your preferences but also allows it to circle back to some of them when prompted.

ALSO READ: OpenAI announces a major ChatGPT privacy update; here’s what you should know

Although helpful, it does end up using a lot of your personal information to help better this feature. To opt out of this, navigate to Settings on your smartphone or laptop app. There, you will be able to locate a toggle option which reads ‘Improve the model for everyone’. Switch it off and you are all set.

However, this step is slightly different if you have an existing account on ChatGPT. Within Settings, find an option that reads ‘Data Controls’. Within this, you can switch off ‘Chat History and Training‘, and make sure you do it for each of the devices you’re using it on.

If you’re a DALL-E user, you might have to go the extra mile and submit a form via OpenAI’s website to remove future training data sets.

Gemini

Google’s Gemini LLM was all the talk last month after the brand introduced a bunch of advancements, including an AI call-scanning option. However, this move, too, gave rise to countless debates about allowing the company to access our day-to-day conversations and data.

If you’re a regular Gemini user, you can always restrict your information by navigating to Activity on the app and select the ‘Disable’ option. This will disable Gemini app activity and simultaneously delete conversation data.

However, note that while future chats may not be reviewed, selected conversations may still remain with the company for up to three years.

Copilot

Microsoft Copilot has a fairly simple way of opting out of data sharing. All you have to do is visit the Microsoft Privacy Dashboard. Based on your usage, you might be required to log in. Next, navigate to control settings where you will find an option that can disable data collection for AI settings.

Best practices to keep in mind while using AI tools and services

Here are a few common practices to follow when using an AI tool or chatbot:

Always use a trusted tool: Today, there are multiple third-party apps claiming to use AI for tasks. Make sure you do your research before downloading or installing a lesser-known AI tool for work or leisure.

Be cautious while uploading a file: Several platforms today support file sharing or image uploading; however, be mindful of where you choose to share your personal photos or sensitive files. With a rise in deepfake technology and cyber-crimes, it is best to avoid uploading photos and videos onto unknown or lesser-known platforms.

Do not rely on AI information: Artificial intelligence does a fairly good job at fetching information for you, but it cannot be used interchangeably with a search engine.

If the data is old, incomplete or factually incorrect, the chatbot can still present to you as an answer to your query. Don’t forget to revisit the answer you generated to ensure you have the right information.

ALSO READ: Google warns users against sharing personal information with Google Gemini; here’s why

Look out for AI-generated communication: Artificial intelligence has made it especially difficult to differentiate between genuine and bot-based communication. Keep an eye out for any messages or emails that may seem out of the ordinary, as these could be a more sophisticated phishing scam.

Do not input personal information into a chatbot: Lastly, like images, do not input information like your address, contact details or confidential financial information into a chatbot. Whether or not you opt out of data sharing, chatbots are still vulnerable to attacks.

AI chatbots have actively helped reduce work and make our lives easier; however, they still require a course of regulations to ensure our data remains safe in the long run. Regardless, keeping up with the above-mentioned hygiene practices in mind could help us go a long way with AI in the future.