Live data from Hacker News

Our approach to age prediction

openai.com

171–180 of 235 posts

Re: Our approach to age prediction

#171
post #156

This could have the unintended consequence of encouraging under-agers to ask more 'adult' questions in order to try to trick it into thinking they're an adult. Analogous to the city that wanted to get rid of rats, so offered a bounty for every dead rat, and to the surprise of nobody except policy makers, the city ended up with more rats, not less. (lesson: they thought they were incentivising less rats, but unintenti…

I am an actual minor and I think I used to watch enough mature content (ie some finance,geopolitics & tech) from youtube where they may have thought I was greater than 18

But when youtube rolled out, I saw this video on taxes simply for tricking the youtube algorithm which had like A LOT of views.

I went to the comments and much of them were teenagers bashing the yt idea and commenting in jest about how yes they got helped in their taxes etc.

I simply don't see how openAI would be any different.

Youtube is still a one in a million though, Nothing else like that exists but there are many chat providers like OpenAI which are actually pretty good nowadays & don't want your id.

Re: Our approach to age prediction

#173

For some reason ChatGPT has suddenly started thinking i'm a teen. Every answer it starts out "Since you are a teen I will..." and prompts me to upload an ID to show my age. I'm 35.

Does OpenAI have an incentive to get age prediction "wrong" so that more people "verify" their ages by uploading an ID or scanning their face, allowing "OpenAI" to collect more demographic data just in time to enable ads?

YES. They all do. Everyone is dripping to get their hands on your biometric data and medical info.

Re: Our approach to age prediction

#174
post #150
post #34

Everyone's saying this is for advertising, but I don't think it is. It's so they can let ChatGPT sext with adults.

It’s worse than that. They want to build “the scream room” from Frank Herbert’s _The Jesus Incident_. There is immense power to wield over someone when you know what they did in the scream room.

Mmm juicy kompromat

Re: Our approach to age prediction

#175
post #96

Earlier quoted context omitted.

Anyone remember the game Leisure Suit Larry? To get the full 18+ experience, you had to answer five trivia questions that only adults should know. But it turns out smart teens who like trivia knew most of them too (and you could just ask mom and dad, they had no clue why you were asking which President appeared on Laugh In).

Also, hilariously, a lot of those questions require a trip to Wikipedia (or a game guide) today. A lot of them reference bits of 1960s/1970s pop culture which are no longer common knowledge. https://allowe.com/games/larry/tips-manuals/lsl1-age-quiz.ht...

This is fun, I asked AI to come up with some modern ones to check someone is over 30. Zune, Friends, early memes, Who Wants to be a Millionaire, etc https://chat.deepseek.com/share/v9d5ckb8gv9rahwetq

Re: Our approach to age prediction

#176
I get why this exists and appreciate the transparency, but it still feels like a slippery middle ground. Age prediction avoids hard ID checks, which is good for privacy, yet it also normalizes behavioral inference about users that can be wrong in subtle ways. I’m supportive of the safety goal, but long term I’m more comfortable with systems that rely on explicit user choice and clear guardrails rather than probabilistic profiling, even if that’s messier to implement

Re: Our approach to age prediction

#177

For some reason ChatGPT has suddenly started thinking i'm a teen. Every answer it starts out "Since you are a teen I will..." and prompts me to upload an ID to show my age. I'm 35.

> simple way to confirm their age and restore their full access with a selfie through Persona, a secure identity-verification service The normalization of identity verification to use internet services is itself a problem. It's described much better than I could by EFF here: https://www.eff.org/deeplinks/2026/01/so-youve-hit-age-gate-...

The EFF are fighting a losing battle:

> we hope we’ll win in getting existing ones overturned and new ones prevented.

All the momentum is in the other direction and not slowing down. There are valid privacy concerns, but, buried in this very article, the EFF admit that it’s possible to do age-gating in a privacy-preserving way:

> it’s possible to only reveal your age information when you use a digital ID. If you’re given that choice, it can be a good privacy-preserving option

If they want to take a realistic approach to age-gating they should be campaigning to make this approach only option.

Re: Our approach to age prediction

#178
post #34

Everyone's saying this is for advertising, but I don't think it is. It's so they can let ChatGPT sext with adults.

Do you expect the data collected for age verification will be completely separate from the advertising apparatus? I would expect the incentives would align for this to enhance their advertising options.

It would be a pretty massive GDPR breach if it wasn't, wouldn't it? All the biometrics is "special category" data which you can't play face and loose with.

Re: Our approach to age prediction

#179

Do not give your biometric & photos of ID's or videos of your face to these companies. Nether to third-parties. The potential failure modes here are very high risk and not worth it. Better to unsubscribe and let them know why.

What is the risk? Anyone who wants a picture of my face can already get one by googling my name and going to my linkedin profile.

Re: Our approach to age prediction

#180

Do not give your biometric & photos of ID's or videos of your face to these companies. Nether to third-parties. The potential failure modes here are very high risk and not worth it. Better to unsubscribe and let them know why.

My face is not private information and probably hundreds of other people's cameras capture pictures/videos of me/my face every day.

I hate age verification as a concept and I wouldn't personally go through it to use chatgpt, but "failure modes here are very high risk" is unnecessarily alarmist.

Post reply on HN