"It's an unfortunate use of advanced technologies, the same ones that are being used to detect and counter misinformation online. It's bad enough that we're having a hard time discerning true from false content, but what if the fake news is tailored specifically for you so that you're the most likely to believe it and consequently share it? Personalised disinformation is the next wave in darkly creative misapplication of technology, and here's how it can work."
"Because your psychological profile is now available, you can be exposed to customised information that matches your personality and increases the chances of the content influencing your future behaviour. For instance, if you're extraverted, you may be more likely to click on an advertisement where people are partying as compared to one where a couple is in an intimate setting-even if the ad is for the same product."
"We all leave digital footprints in our online lives. A post you shared, a comment you liked, a video you bookmarked, a photo you uploaded, a song you listened-it all adds up. This data can be used to reliably predict the kind of person you are-essentially, your personality. (Side note: The ethics around collecting user data without permission or with surreptitious permission are very murky and beyond the scope of this book.)"
AI can create deepfakes and misleading content at scale, infesting the information landscape with doctored stories. Personalised disinformation tailors fake news to individual psychological profiles to increase the likelihood of belief and sharing. Everyday digital traces—posts, likes, bookmarks, uploads, and listening habits—accumulate into data that can reliably predict personality. Those profiles enable customised content matched to personality traits, thereby raising the chance of influencing future behaviour. Example differences in ad responsiveness illustrate the tactic. Ethical concerns arise from surreptitious data collection, and awareness of data use can help people resist targeted persuasion.
Read at Psychology Today
Unable to calculate read time
Collection
[
|
...
]