Will My Private Questions to ChatGPT End Up Public?

Since its unveiling in December 2023, ChatGPT has become the face of new, revolutionary technology. The cutting-edge AI has numerous benefits for personal and corporate users. However, bright red flags flame around the chatbot as its glitches still need to be ironed out.

The technology has gained rapid popularity, and over 1 billion people have used its services. While the powerful AI shows hope and promise, with great power comes great responsibility. It remains to be seen how OpenAI, the creator of ChatGPT, will meet and conquer its challenges, which sparks some questions for users.

Questions private users may have include:

• How does ChatGPT work?
• Is the sensitive and confidential data you plug into ChatGPT saved?
• Will your private questions to ChatGPT end up public? 
• How do you protect your online security while using ChatGPT?

Let’s take a look at the answers. 

How does ChatGPT work?

Open AI’s ChatGPT is a conversational, language-based AI that relies on Natural Language Processing (NLP) to communicate with users. Human feedback and input allow the AI to constantly learn and evolve, and allow it to give quasi-human-like responses to user questions.

Important facts about ChatGPT and how it works include:

• OpenAI unveiled its launch of AI on November 30, 2022.

• The ChatGPT input and training come from over 300 billion words and 570 gigabytes of text data.

• ChatGPT can remember past “conversations,” explain its answers with detail when asked, and will apologize when called out for incorrect answers.

• With input, ChatGPT can create written content — including fiction and personalized letters. A user must still edit and revise the content, but the basic story and outline can be generated.

• The chatbot also acts as a language translator, easily translating text from one language to another.

• ChatGPT can provide vital customer service tools for 24/7 support.

• Rather than acting as a search engine to gather data, the chatbot relies on the human input it has received to give answers.

As ChatGPT receives more human input, its database grows. The version of ChatGPT available now will likely look vastly different from the ChatGPT models of the future.  

Where does ChatGPT get its data?

The GPT in ChatGPT stands for Generative Pre-trained Transformer, which is the aspect of the program’s architecture that allows for conversational text creation, question answering, and translation. 

The AI’s responses to users are meant to represent human responses, rather than robotic, obviously computer-generated answers. ChatGPT gets its data from its human engineers who input countless books, journals, articles, and blog posts into its database. These texts are curated to ensure diversity in the chatbot’s archives.

ChatGPT operates differently than a search engine like Google, as its curated knowledge has been carefully selected, and it’s not connected to the Internet. 

Does ChatGPT save your sensitive data?

The good news is, OpenAI has a privacy policy in place that keeps any user data information safe and secure. The sensitive and confidential data it collects from you includes social media accounts, payment information, and content you’ve uploaded. The privacy policy is parallel to that of many online-based businesses.

Nonetheless, OpenAI uses cookies and information about the device you’re using to connect to ChatGPT services. The company also stores your chatbot conversation data — which may or may not include sensitive or identifying information. There’s nothing nefarious behind this storage, and it’s used to improve ChatGPT’s abilities.

Will your private questions to ChatGPT end up public? 

If you ask ChatGPT a question that contains sensitive, confidential, or identifying information, you don’t have to worry about public exposure. OpenAI utilizes data encryption to keep your private questions away from prying eyes. ChatGPT now includes a private mode, which will prevent others from accessing your chat history.

However, if you remain in public mode, questions from “Why did I get dumped?” to “Is there a meaning to my street name (Snuffaluffagus Way)?”  can be accessed by other users in your area. And specific questions may reveal your identifying information. Your question will remain in the ChatGPT database, and you’ll have no way to erase it from the public domain.

Is ChatGPT a threat to your privacy? 

Although OpenAI takes the security of its users seriously, there are still issues that may arise. For example, privacy concerns prompted Italy to outright ban ChatGPT. On March 31, 2023, the Italian data-protection authority announced the ban and investigation of ChatGPT.

The watchdog regulator cited OpenAI’s sourcing of information from blog posts and social media accounts as troublesome. ChatGPT doesn’t differentiate between personal and public data when “learning” and doesn’t ask for permission before using these sources. The concern and caution over the chatbot’s loose regulations and rapid growth are valid. 

Even in the U.S., technology journalists have sounded the privacy breach alarm. Ars Technica calls ChatGPT “a data privacy nightmare” as the bot contains 300 billion words from various, unsuspecting sources. Thus, although privacy and security parameters are in place for the chatbot users, unsuspecting data sources (which could be any of us) never gave permission for their words to be used.

How to protect yourself when using ChatGPT

Users of ChatGPT have more control over their data protection than those who have been sourced by the AI. Currently, no one has control over giving or denying access to ChatGPT’s database. 

For example, the terrible, break-up poetry blog you published for your 9th grade crush and then tried to push to the recess of your memory? If it still exists online, it may have found its way into the knowledge base of ChatGPT. 

However, when you use the chatbot, there are ways to protect your sensitive information when giving private questions to ChatGPT and steer clear of hackers. Here are a few things to watch out for, and ways to protect yourself:

• Malware and viruses: Scammers could pose as known entities on a chat, and use ChatGPT to prompt you to download malware or other computer viruses.

• Phishing scams: Online criminals may use ChatGPT to create phony websites that look almost identical to original, legit websites. Make sure you check the URL and IP address before clicking on any links.

• Utilize the privacy mode: When you’re using ChatGPT, take advantage of the privacy mode feature. This way, random strangers won’t have access to your chat and question histories.

• Confirm identities: If someone is using ChatGPT to connect with you online and pose as someone else, their language may seem a bit off or clunky. The chatbot is still working on conversational glitches. However, this may be a way for a scammer to hide behind their purported identity with legitimate facts about the person they’re impersonating. Online criminals may use ChatGPT to try to access your money or confidential information by posing as a credible person or organization. 

• Implement ChatGPT policies and procedures for your business: Businesses should limit employee’s access to on-the-clock ChatGPT use and restrict its use to tasks which benefit the organization. The chatbot has numerous benefits for businesses including the undertaking of customer support and research work. Nonetheless, confidential and sensitive business data should be prohibited from sharing in ChatGPT. Implementing policies can prevent security breaches while still enjoying the upside of AI.

Learn more to protect your online safety

At What is My IP Address, we care about your online safety. By answering questions like, “Will my private questions to ChatGPT end up public,” we hope to help you remain informed about the benefits and drawbacks of the latest cutting-edge technologies.

We offer various downloadable tools to support your online security measures, and keep you safe. Check out our blog for the latest on computer security threats and protection tips.