Raven finds a printed Sky Sorter report in the recycling bin.
Next to each snapshot is a note: who took it, when, and the exact spot of their backyard telescope on Earth.
"Anyone who finds this knows where Rocket's family lives," she says quietly.
Rocket turns pale. "I just wanted credit for my snapshots. I did not think about the location."
Nova scans the page. "The snapshots are fine to share," she says. "The location was a leak."
"So the data was useful," Raven says, "but part of it escaped to the wrong place."
AI systems are built from data, and some data is about people.
A data privacy risk of generative AI is leaking or misusing personal information.
That includes health information, location information and biometric information.
Biometric information comes from your body, like your face, fingerprint or voice.
Leaked information can be used in ways nobody agreed to. It can be hard to take back.
| Kind of personal data | Story example |
|---|---|
| Health | Raven's checkup results |
| Location | Where Rocket's backyard telescope sits |
| Biometric | A close-up photo of someone's face or a voice recording |
| Identity | A full name with a birthday |
| Statement | True or false? |
|---|---|
| Leaking personal data is a data privacy risk of generative AI. | ? |
| Location data cannot reveal anything private. | ? |
| Biometric data comes from a person's body. | ? |
| Once private data leaks, it is always easy to take back. | ? |
The safest data is data you never give away.
Never type personal information into any app or website, including a chatbot.
Ask a grown-up before using any AI tool outside this course.
Careful thinking. Tomorrow in the Explorer Lab you test how locks and codes protect information.