Following Discord's suit, OpenAI will also predict your age based on your usage to know if it should ID you at the digital door
With OpenAI implementing more safety settings in ChatGPT and giving it an adult mode at some point this year (which was actually supposed to launch last December), it is now rolling out an age verification program. If, however, you regularly use ChatGPT, there's a chance it can verify you with information it already has.
As shown in a recent ChatGPT help article, "It looks at different signals linked to your account. For example, it may look at general topics you talk about or the times of day you use ChatGPT." It does note, "No system is perfect. Sometimes we may get it wrong."
If you can't or won't verify your account, ChatGPT will implement safety features that attempt to reduce content such as:
- Graphic violence or gore
- Viral challenges that could push risky or harmful behavior
- Sexual, romantic, or violent role play
- Content that promotes extreme beauty standards, unhealthy dieting, or body shaming
If you do get your account verified, it will bypass these safety settings, which should lay the groundwork for more adult-themed updates later this year. The post clarifies, "In Italy, you must complete this verification within 60 days of being prompted, or certain features will be disabled." It does not name which features will be dropped, though.
Even if you entered your correct birthday when setting up your ChatGPT account, it will still need to verify you. If, however, it cannot successfully log your age based on your past behaviour, you can choose to do so via Persona, a third-party company OpenAI is partnered with.
OpenAI claims it "does not see what you share with Persona. Persona deletes it after verification. OpenAI only learns that you verified you’re 18 or older and information about your age (for example, a date of birth), not the ID itself."
Persona can ask you for both a government ID and a live selfie using a webcam or your phone. Failing to verify your account doesn't block you from using ChatGPT, but it does put in place those safety measures.
Just this week, Discord confirmed a very similar process. It too will start to verify users based on past behaviour, and will require an ID or live photo if it fails to do so. Still, despite measures in place to not save information, it's still a tad alarming the information these websites have on you if they can confidently guess your age based on your behaviour.
I don't know which outcome would feel worse: a site being able to guess my age due to the information I've given it, or it guessing I'm a teenager based on my usecases. Either way, I feel like they know too much.