OpenAI is said to be rolling out an age-prediction system for ChatGPT. This is not just for fun and games – it’s trying to guess whether a user is under 18 and then quietly limits what that account can see or ask.
The company apparently says that it knows it won’t always get things right, but this of course raises tons of questions about privacy, accuracy, and who gets to decide what’s appropriate out there in the internet space.
Tech expert Omar Gallaga has been following these developments and joined Texas Standard to discuss.
Highlights from this segment:
– Many have already told ChatGPT their age, which the system will remember. But the program is also rolling out an age verification system. If an adult is flagged as a minor, they’ll then need to go through age verification.
– The program will use things like a user’s time of day login and past interactions and conversations to determine the age of a user. If flagged as a minor, the program would begin filtering things like graphic violence, depictions of self-harm, risky viral challenges, body shaming and other potentially harmful content.
– Because age verification would require a live selfie and government-issued ID, concerns over privacy are being raised.
