Why Do We Sometimes Trust AI More Than Websites?
From GDPR to AI – A Surprising Shift
It has only been a few years since the General Data Protection Regulation (GDPR) came into force, with the aim of bringing order to the storage, processing, and sharing of all our personal data. Going forward, we had to explicitly allow every website to store our IP address or track our visit using a statistics tool. The result was not just a flood of warning letters from dubious law firms. It also created almost a sense of paranoia among us all that it could have terrible consequences if we had to provide our email address somewhere on a website, as it could somehow be misused.
And now, a few years later, we are allowing some AI—whose inner workings nobody really understands—full access to our email inboxes. We allow Copilot to read our emails, entrust ChatGPT with personal problems in the hope of helpful answers, and let Gemini see everything that our smartphone camera captures. And all of this directly and traceably linked to us as individuals, rather than „just“ to an IP address or perhaps an anonymous inbox.
Why We So Readily Trust AI
Now, you would think that with everything we entrust to AI, alarm bells would start ringing and we would carefully consider what we share with various AI tools. But the exact opposite is the case. We are delighted when we ask our „personal AI assistant“ a question and it already knows exactly when we are flying on vacation, what is on our shopping list, when our next doctor’s appointment is, and what the temperature is in our living room right now.
Between Convenience and Data Awareness
I am certainly not immune to this myself; I use AI in both private and professional contexts. I too provide AI with emails or texts, for example, to have them summarized. I also let AI visually enhance every single one of my vacation photos. And yes, I might even ask an AI questions that I would probably be better off asking my doctor. However, I don’t do this indiscriminately, but rather with selected AI tools from well-known manufacturers whose products I have been using for decades and who probably knew almost everything (relevant) about me even before the AI hype.
Indeed, thanks to this AI hype, new AI tools feel like they are coming onto the market every day, promising even faster, even better, and even more amazing results. And of course, one is tempted to try out these tools (which are usually free at first). And we quickly relapse into the thoughtless steps of uploading something or asking questions that reveal far more about us personally than we realize.
Talking to AI Is Not a Private Conversation
Speaking of asking questions, the mere fact that interacting with most AI tools feels like a conversation seems to significantly lower our inhibition threshold regarding our personal data. After all, you are talking to your „personal assistant“ and not, say, uploading something via some ominous form on a website where it remains on the internet forever and outside your control.
But we shouldn’t let ourselves be fooled by that. The friendly chat with ChatGPT doesn’t change the fact that everything we disclose there is stored by the AI and remains linked to us. The fact that Copilot kindly reminds me to pay Max Mustermann’s invoice doesn’t change the fact that Microsoft might now know I am doing business with Max Mustermann.
Using AI Consciously Instead of Trusting Blindly
AI is here, and it has become an indispensable part of our digital everyday life. But all the convenient benefits and new possibilities that AI offers us should not strip us of a healthy skepticism and a conscious approach to our own data. Whether it’s AI or some websites from ten years ago… our personal data and information should remain what they are: personal. And only we personally decide to whom and where we make them available.
Article by HeiReS employee Armin Reuter – may contain individual opinions
