These days, it seems like you can’t go anywhere without hearing someone mention artificial intelligence (AI). From businesses leveraging AI capabilities to individuals playing around with ChatGPT, it’s safe to say that the AI takeover is officially here (hopefully not in the Terminator way). But, is ChatGPT safe?
Whether you’re getting help with your homework, generating art, or using one of the many other GPTs, here’s what you need to know.
What is ChatGPT?
You’ve probably heard of ChatGPT, but if you aren’t quite sure what it is, we’ve got you covered. Without getting too into the weeds, ChatGPT is an AI chatbot that uses a subset of AI known as generative AI (GenAI) to generate anything from human-like conversations to images and beyond.
Who's Calling Me?
Search any phone number to learn more about the owner!
The “GPT” in ChatGPT stands for Generative Pre-trained Transformer, which is simply the language model framework that GenAI is built around. It can all sound a bit like sci-fi-talk if you’re new to the concept, but it’s essentially a tool that is able to use natural language processing and machine learning to interact with people in a way that’s similar to a human…if that human had the entire internet in their brain.
What is ChatGPT Used For?
Because ChatGPT is an incredibly powerful tool that is only getting better with additional training, the uses are widespread when it comes to generating personalized responses about, well, pretty much anything. For consumers using ChatGPT, here are some things it’s most often used for:
- Answering questions with in-depth information
- Assisting with writing based on a given prompt
- Summarizing documents into key points
- Writing or debugging code
- Creating images/art based on specific prompts
- Creating personalized recommendations (i.e., “create a weight-loss workout plan” or “best Italian restaurants in LA”)
The list of potential uses can go on nearly endlessly, but the above points should give you a pretty good grasp of what many people use it for.
Is ChatGPT Safe?
If you’re wondering whether ChatGPT is generally safe to use, the answer is yes, it’s safe. However, there are still some potential risks you’ll want to be aware of when it comes to ChatGPT.
Phishing and Smishing
Pretty much anytime scams are being discussed, phishing is going to be mentioned. For those in need of a quick refresher, phishing and its text-based sibling, smishing, are scams where victims receive an email or text message that appears to be from a legitimate source containing a malicious link. This link then either takes victims to a cloned website or prompts malware to be downloaded onto their device.
So, how does ChatGPT come into play? Classically, a telltale red flag that you were receiving a phishing or smishing message was poor grammar or odd language usage. That’s because a large portion of scammers are located outside the U.S. in non-English-speaking countries. Now, with the help of ChatGPT, messages are not only nice and polished, but they can even be created to match the tone of a company they are pretending to be, making it extra difficult to discern real from fake.
Catfishing
If you’re unfamiliar with catfishing, we’ve covered it in great detail (learn more here), but it’s essentially the use of a fake online identity to trick others for malicious purposes, such as money, private information, or even blackmail material.
This is almost always done with fake photos, usually stolen from a specific person’s social media account. In recent years, reverse image searches and people-finder tools have made it easier to catch catfishers using stolen photos, which has led these scammers to ChatGPT’s image-generating abilities. (Even more frightening, scammers have also begun to use AI voice generation to really strengthen their gambits.)
Because ChatGPT can make essentially photo-realistic images, scammers have begun using it to make whole portfolios of (often very attractive) fake people, all for the purpose of catfishing unsuspecting victims into romance scams.
While these photos won’t appear in searches like stolen photos, AI is still wrinkling out some issues when it comes to the details, so look for things in the photos that look off, like weird hands or strange words in the background.
Spreading of Misinformation
AI is great at quickly providing users with a detailed answer. However, there’s currently the slight issue that AI has the information wrong, and as more people use that wrong information on the internet, it reinforces AI’s belief that it’s giving people good answers.
The result — a snowball of misinformation. While it can be hard to fact-check some of AI’s misinformation, especially as more people blindly rely on what AI tells them and thus inundate the internet with falsities, one thing you can do is directly check any sources ChatGPT provides.
When you’ve received an answer to an inquiry on ChatGPT, ask it to provide the sources it used on the information, and then actually take the time to check the linked source for the matching facts, as ChatGPT has been known to essentially make things up and randomly attribute sources. So long as you take this extra step, you should be safe with using ChatGPT to help you get the facts straight.
Malware Creation
Malicious software, commonly known as malware, is a common tool used by scammers to compromise people’s computers in order to access any sensitive private information contained within. In order to create or improve malware, cybercriminals typically need to have a certain level of programming knowledge to create something that works. With the use of ChatGPT, the required level of know-how is substantially reduced.
While ChatGPT has general safeguards in place to try to prevent it from being used for illegal means, that hasn’t stopped people from finding loopholes that allow for the creation of sophisticated malware. While this isn’t a safety concern if you’re worried about using ChatGPT, it’s worth noting a general safety issue. Fortunately, avoidance methods aren’t all that different from traditional malware, in that you don’t want to click or download anything unless you’re 100% confident in the source.
Fake ChatGPT Clones
Scammers love to clone real sites and apps in the hopes that people will unknowingly visit/download them. ChatGPT is no different, with plenty of knock-offs already around hoping to steal people’s private information. While there are other chatbot AIs outside of ChatGPT (think of ChatGPT more like a brand that uses AI, in that there are other GPTs that aim to do the same things), scammers will use the term “ChatGPT” to gain clicks, since people are most familiar with that name. When using ChatGPT, make sure it’s the correct website or app created and owned by OpenAI.
Confidentiality
Our final safety concern when it comes to ChatGPT is the confidentiality of both the information you provide and the information it spits out. As it stands, ChatGPT openly states that users should not enter confidential information, as there is no expectation of confidentiality. In simple terms, if you have information or source code that you don’t want potentially exposed, don’t enter it into ChatGPT.
AI Can’t Fake Real People
If you’ve made it this far (or just scrolled down to find the answer), you can feel confident in using ChatGPT. Although ChatGPT has created some safety concerns, it’s safe to use and is going to be a powerful tool for years to come. Because most of the direct concerns have to do with scammers using ChatGPT as a tool to help boost their scams and create fake profiles, remember that people search tools like Spokeo can still validate if…
- A person actually exists, and
- If the information that person is telling you lines up with their actual information.
Both of these are invaluable tools for navigating the digital world, no matter if the content you’re interacting with was created by ChatGPT or a real person.
Cyrus Grant is a writer from Southern California with a background in law and dispute resolution. When he isn’t writing he can be found deep-diving into the latest technology trends or simply spending time at the beach.