By Ricky Browning · Browning PC, Valdosta, GA
We usually write about AI in the context of saving you time, automating busywork, or helping your small business run smoother. But AI has another side that doesn't get talked about as much — and a recent case right here in Georgia shows it off in a powerful way.
A man is now spending the rest of his life in prison, and the chain of events that put him there started with the one thing he never expected to turn him in: ChatGPT.
In Eatonton, Georgia — about two hours north of us here in Valdosta — 38-year-old Corey Hickey was living with a roommate and their young daughter. Over a period of several months, Hickey repeatedly abused that child, who was just seven years old.
At some point, Hickey uploaded two images of the victim to ChatGPT. We don't know exactly why — but it didn't matter. OpenAI's systems flagged the images as child sexual abuse material (CSAM) almost immediately and forwarded them to the National Center for Missing & Exploited Children (NCMEC).
From there, NCMEC generated a "cyber tip" that landed on the desk of the Georgia Bureau of Investigation's Child Exploitation and Computer Crimes Unit (CEACC). The GBI arrested Hickey in late October 2025. When investigators searched his devices, they found more than 4,800 images and videos of child abuse material on his phone, plus over 400 more on his Google Drive.
On May 21, 2026, a jury convicted Hickey on all 31 counts — including three counts of rape of a child under 10, four counts of aggravated child molestation, three counts of child molestation, and 21 counts of sexual exploitation of children.
He was sentenced to multiple life terms (two without the possibility of parole) plus an additional 220 years — believed to be the longest sentence ever handed down in the Ocmulgee Judicial Circuit.
This part surprises a lot of people, so let's break it down in plain English. When you upload an image to a major platform — ChatGPT, Google Photos, Facebook, iCloud, Dropbox, and many others — that image is often scanned automatically against known databases of illegal material.
The technology behind this is usually a system called "hash matching" (the best-known version is Microsoft's PhotoDNA). Here's the simple version:
OpenAI has publicly committed to detecting, blocking, and reporting CSAM on its platforms, and it works directly with NCMEC and groups like Thorn to do it. In other words, this wasn't a fluke. The system worked exactly the way it was designed to.
💡 The bottom line: A predator handed his own evidence to an AI tool, and that tool reported him. A seven-year-old's abuse was stopped, and the man who did it will never walk free again.
Stories like this raise a fair question from our customers: "If ChatGPT is scanning images for that, is it reading everything I type and upload too?" It's a reasonable thing to wonder, so let's be straight with you.
This is the real takeaway for everyday users. The things you type into ChatGPT, Gemini, Copilot, and similar tools are not like a private diary. Depending on your settings, your conversations can be:
For the overwhelming majority of people, that's nothing to worry about. But it's a good reminder of a rule we give all our clients: don't put anything into an AI chatbot that you wouldn't be comfortable having stored, reviewed, or seen by someone else. That means no passwords, no Social Security numbers, no sensitive client data, no banking details.
We talk a lot about the risks of AI on this blog — from AI scams to the way screens affect our kids. But it's only fair to highlight the good, too. This is exactly the kind of thing the technology should be doing: protecting the most vulnerable among us and giving law enforcement the tools to act fast.
The same hash-matching and AI detection systems that caught Hickey are quietly protecting children every single day across every major platform your family uses.
You can't catch every predator, but you can make your own home a much harder target. A few things we recommend to every family in South Georgia:
💡 If you ever encounter child exploitation material online, report it to NCMEC's CyberTipline at report.cybertip.org or call 1-800-843-5678.
If you'd like help setting up parental controls, locking down your family's devices, or just want a plain-English walkthrough of how to use AI tools safely at home or in your business, that's exactly what we do. No contracts, no judgment — just honest, local help.
According to the article, Corey Hickey of Eatonton, Georgia uploaded two images of his victim to ChatGPT. OpenAI's systems flagged them as child sexual abuse material almost immediately and forwarded them to the National Center for Missing & Exploited Children, which generated a cyber tip that led the Georgia Bureau of Investigation to arrest him in late October 2025.
The post explains that images uploaded to major platforms like ChatGPT are often scanned automatically against known databases of illegal material using hash matching, such as Microsoft's PhotoDNA. If a fingerprint matches known abuse material, U.S. law requires the platform to report it to NCMEC. Newer AI classifiers can also flag brand-new abuse images not yet in any database.
The article says AI chats are not as private as people think. Depending on your settings, conversations can be stored on company servers, reviewed by human staff, used to train future AI, and scanned for illegal content and reported to authorities when the law requires. Browning PC advises never entering passwords, Social Security numbers, sensitive client data, or banking details into a chatbot.
Browning PC, in Valdosta serving South Georgia, helps families set up parental controls, lock down their devices, and get a plain-English walkthrough of using AI tools safely at home or in business. The post describes honest, local help with no contracts and no judgment. You can call or text 229-561-1674 or email ricky@browningpc.com.
📞 Call or text us at 229-561-1674
✉️ ricky@browningpc.com
📅 Book a remote or onsite appointment
Sources: 11Alive, The Union-Recorder, Georgia Bureau of Investigation, and OpenAI.