Harmless Help or Digital Weapon? The Reality of Modern AI

ChatGPT. Copilot. Gemini. Perplexity. Take your pick. The truth is, in the past 5 to 10 years, the use of artificial intelligence has skyrocketed. Enter chatgpt.com into your web browser and you can ask it to write an essay or brainstorm ideas for a project with just a couple of words and the enter key. All harmless, right? Well, other people don’t exactly use AI for those same reasons. While you might be using it to check the quality of your email, others may be using it to check your location. While you might use it to find information about who’s running for your local council, others might be using it to spread political misinformation. You could simply upload a photo of yourself onto any social media platform, and some people would use AI to alter your image, and then upload that, without your consent. The weaponisation of AI, particularly against specific people in society, is one of the most significant issues impacting our generation.

Imagine waking up to find a video of Donald Trump doing the “Big Guy” trend. Sadly, this isn’t just an imagination for me. I had just finished my breakfast and wanted to relax and watch something before going out. And what do I find? A video. Of Donald Trump. Doing the “Big Guy” trend. Even though it was realistic, I knew it was fake. But what if the video was something more believable? Like him signing a bill related to immigration laws. Would I have had a harder time trying to work out whether it was real? In today’s world, anyone can log onto an AI platform and create a video that can be forwarded faster than it can be fact-checked. And if it’s a political figure? People often react and share their opinion long before they try to work out whether it is to be trusted. And why would they? The purpose of these videos is to spread misinformation, to turn society against a politician. It’s done to influence elections and divide communities, to create panic and anger. It’s done to damage trust in leaders, when these very leaders haven’t even said a word. And there have been major public instances of this. During Donald Trump’s campaign, he shared an AI-generated image of an endorsement from a well-known music figure; Taylor Swift. And even though it was obviously fake, it begs the question of how easily it was used in a political setting and the stark resemblance to the pop star. And that’s the danger; if a fake video can look real enough to influence millions, how does the truth even stand a chance?

Is there an image of you anywhere on the internet? Well, would you be alright with someone taking that image, editing it, and then uploading that image to the same, global internet? An image you would never, never post. These images can harm someone’s reputation, even if it isn’t real, following them for life. AI has gotten to the point where it is nearly impossible to draw the line between what is real and what isn’t. And these images? Some of them may be used for defamation. Some are used as blackmail. Some AI can even create non-consensual sexual images that can be easily published wherever the creator wants. All without consent. 98% of explicit videos are deepfakes, and 99% of these are women. And this doesn’t only happen to adults. In September 2024, 18-year-old William Hamish Yeates was charged with creating sexually explicit deepfakes of a teenage girl and sharing them online. Imagine that was your best friend. Your sister. Your daughter. You. And in a world where anyone can create these images, the line between truth and fabrication disappears, and anyone can become a target.

With a single photo, AI can collect dozens of pieces of data about you. Not because you told it, but because of the signs, the background, even the lighting in your photo. And it won’t stop at just one post. If you have a social media account, people can use AI to check every single picture you have posted. This trail leads to a routine that anyone can trace. The ground where you play basketball, the restaurant near school you go to eat at. And to people who actively want to find out where you are? AI can serve it to them on a silver platter. At the beginning of December last year, Brett Dadig was sentenced to around 70 years in prison on 14 counts of cyber-stalking, interstate stalking and interstate threats related to 11 victims in Pennsylvania as well as New York, Ohio, Iowa and Florida. 14 counts of stalking. Dadig weaponized AI and made it possible to stalk these unsuspecting people. Anyone could use AI for these same reasons. Everyday photos turn into clues, your routine being the red string AI uses to join them together. The stalker looks at your life like a detective looking for the next murder. The next murder that will happen long before you see it.

The misuse of AI isn’t some foreign problem we might have to deal with in the next 10 years. It’s a problem that is happening right now. And the key point is this; AI itself isn’t bad, but instead how it is used by members of society. Though some may argue that AI is just a tool, or that these cases don’t happen, it can be easily disproved with evidence. People are using AI for the wrong reasons.  So how do we fix this problem? A start could be to implement much stricter guidelines and laws around how AI can be used. Less workarounds, less grey space and more clear rules. Because if we don’t put up these protections in now? The weaponisation of AI won’t only affect our generation, but every generation after us as well. So, take your pick. Perplexity. Gemini. Copilot. ChatGPT. Just beware; anyone can use these services. For anything.