I’m not an AI user for a long list of reasons. I’m not going to get into them now, though, and I don’t want any readers who do use AI chatbots to feel judged by my avoidance of them. You do you, but if you are going to ask them for help, I recommend that you use a VPN while you do it. Your friendly little robot pal is actually absorbing a lot of your data, and that’s something for you to be wary of if you value your privacy.
What got me thinking about this is the social media trend of people asking ChatGPT to make a caricature of them at work. I already know that, if I did it, it would make a picture of a radiantly beautiful human being writing world-changing journalism, so I don’t need to. Despite my anti-AI stance, I have to admit that seeing these pictures of people I care about elicited at least a small smile from me. Though the cynic in me thought, “This seems like a ploy to get data out of people.”
I thought the same thing last year when people were using AI to make images of themselves as action figures with accessories from their lives. Wanting to learn more about the data risk here, I asked ChatGPT itself – the first time I have ever used the service. My first instinct, of course, was to ask whether or not it was selling user data, and it said, definitively, “No – I don’t sell your data.”
That might sound somewhat reassuring, but then I asked it what data it accumulates on its users, and it gave me this big list:
- Email address
- Username
- Subscription plan
- Messages you send and responses generated
- Features used
- Timestamps
- Device/browser type
- IP address (for security and fraud prevention)
- Files, images, or documents you choose to share
So even if the data isn’t being sold, there is a lot of information about you that is tied to your username and email address. The data it uses to make your caricature will stay with it forever. If OpenAI has a data breach, as many giant tech companies have in the past, that’s an enormous amount of potentially very personal data that can be very easily tied to you. It is also advantageous for ChatGPT to have this information because A) you’ll be feeding its machine learning and informing its future replies to people, and B) when it rolls out ads, these will doubtlessly be tailored based on the data it has on you to make you more likely to click.
I also thought to myself, “Well, you little robot rascal, you may not be selling our data right now, but what if you change your mind in the future?” so I asked it whether it would legally be in a position to sell our data in the future, and its response rubbed me the wrong way.
“I’m a system that generates responses. I don’t own, store, or make legal decisions about user data.” It made me feel like I was interviewing a politician, gesturing to a freshly signed bill on the table and saying, “Aren’t you afraid that this legislation will endanger people’s lives?” and they respond, “No, I’m not afraid of that, it’s just a piece of paper.”
In fairness, it did go on to say, “In theory, any company can change its privacy policy in the future,” and then explained that customers would need to be notified of this change of policy, but likely, all that would be is “We updated our privacy policy, please accept to continue using”, which millions would click without a second thought.
Either way, I think it’s inevitable that every piece of data you give to ChatGPT will end up being sold or stolen eventually and used to manipulate you one way or another. So what’s my advice to those of you who want to keep using chatbots, but want to protect your data?
Well, first sign up for one of the best VPN services (like NordVPN, Proton VPN, Surfshark, CyberGhost, or ExpressVPN) and then only use ChatGPT or other chatbots while the VPNs are on. Nord is our top recommendation, but they’re all good. Then, if you want it to give you random pieces of information or advice, you can do so without it building a huge profile on you. Of course, there’s also the option of just not using it.
VPN newsVPN news#VPN #you039re #ChatGPT #chatbots1771107211
