How AI Uses Our Data (And What it Wants Most)
#HOWTO | AI creates a massive demand for consumer data, but until we understand what information is used, and for what purpose, it's difficult to get a handle on how to protect our privacy.
I recently created a presentation explaining the very specific ways that AI uses data and how AI innovation changes the risk profile of information we previously thought of “low risk.” The link below will take you to the slides outlining what data is most in-demand now for AI, and why we need to think of privacy risk differently.
The Yearbook Example
In, say 1995, you would expect:
Photos were seen by a few hundred people (at most)
To observe an image would require the physical possession of a book
Images were generally low-resolution and difficult to reproduce
There was, of course, no way to automate identification
Now in 2025, the landscape has changed:
Photos can be digitized and processed with facial recognition technology
Images can be combined with a name to create a persistent identifier
High-quality images are valuable to train age-progression algorithms
Images can be matched to social media profiles to create behavioral overview
Emotional patterns can be analyzed through images
Photos can be processed to analyze physical characteristics
All photos in a yearbook can be useful as reference data for deepfake generation
Images can be sucked into large training datasets without owners being able to track their personal photos
If we know the risks, we can address them, so read on for more details and steps for staying savvy and safe.