YouTube Officially Rolls Out AI Age Verification, What You Need to Know

Does Google’s AI know if you lie about your age?

YouTube rolled out AI age verification on August 13, 2025

Have you ever input a fake birthday when signing up for an online account? I have. It used to be standard internet safety advice, and it’s the reason I have a hard time logging back into my Neopets account whenever I forget my password. And it makes sense to lie—birthdays are sensitive information.

But technically speaking, you’re not allowed to even have a YouTube (or Google) account if you’re under 13. A parent or guardian has to “enable” YouTube for kids younger than that. This is a common rule for social media terms of service due to COPPA, a U.S. law that prevents anyone on the internet from requesting or tracking kids’ personal information.

X requires a parent’s government ID to create a kids’ account, and Instagram makes accounts private by default for anyone under the age of 16. But neither of these policies is perfect—both can be avoided by inputting a fake birthday.

That’s why YouTube is pushing this new AI verification. It’s actually a slightly better policy than other services are rolling out right now. For example, the U.K. Online Safety Act recently forced Discord to ID British users who want to access 18+ servers or use other restricted features. The ID system is infamously flawed and can be easily bypassed if you have a copy of Death Stranding.

At least with an AI, there’s a chance that you can continue using the product without having to upload an ID or a weird video of yourself opening and closing your mouth. YouTube’s AI allegedly works by tracking a “variety of signals” like “the types of videos a user is searching for, the categories of videos they have watched, or the longevity of the account.” You need to be 13 to make a Google account, so hypothetically speaking, accounts older than five years should be good to go.

Unfortunately, there’s a big gray area there. There’s no guarantee YouTube is actually doing that math, and the algorithm might rely on your watch habits instead. It’d be great if its AI was 100% accurate, but YouTube viewers are saying otherwise. Shortly after the program began, one Reddit user reported that YouTube asked for age verification on a work account that they had mostly used to research  “international relations and military affairs.” Meanwhile, multiple teenagers commented on their post that they hadn’t been flagged at all.

The flagged user also claimed that their account was a decade old. So YouTube’s algorithm saw an account that spent the last ten years watching military documentaries and decided it must have been created by someone younger than eight. I guess that could be … possible? But not exactly the smartest move by what’s supposed to be a cutting-edge AI.

Ironically, the video that YouTube blocked that user from watching was about a symbolic anti-AI protest movement centered around the 90s tech icon, Clippy. Could just be a coincidence, but it’s not a good look when a platform is age-restricting videos that criticize it. Doing my due diligence here: I tested that video on a logged-out browser and had no problem accessing it. I also tried watching adult-rated videos from Helluva Boss and had no problems there either. It’s not currently clear when YouTube decides to restrict a video, and when it doesn’t.

Another problem is that going forward, the new policy becomes kind of a catch-22. In order to watch adult content, you need to verify your status as an adult. But in order to verify your status as an adult, you need to watch adult content. Do you see the problem?

But if you’re worried because you’re a 35-year-old adult who listens to the Steven Universe soundtrack a lot, don’t be. You probably have just as much a chance of getting flagged as someone who searches up nothing but videos from accounting influencers about proper tax filing. YouTube recently had to assure viewers that it’s not discriminating against people who watch a lot of Roblox or Minecraft content, although that just makes its policy more confusing. What activity is it exactly monitoring? It’s impossible to know, because AI agents like YouTube’s are often a black box.

What to do if YouTube incorrectly labels you as a child

You’ll know if YouTube failed to correctly guess your age: At some point while using the platform, you’ll see a pop-up asking you to “Take a few minutes to verify your age.” Here’s what you should do if this happens:

1. Don’t panic. If the AI fails, you’ll have the option to verify your age another way. And if you choose not to verify at all, you’ll still have access to all non-restricted YouTube videos. Some websites and AI assistants might claim that your account will get deleted, but these claims are false and based on an incorrect interpretation of unrelated policies regarding Google accounts for young children. Feel free to click “Not now” and watch a different video instead.
2. If you choose “Verify age,” you’ll have three options for manual verification: government ID, selfie, and credit card. Most people will probably opt for the selfie option, although it’s worth noting that a credit card is easiest to immediately replace if compromised.
3. Once you submit one of these three forms of proof, you should have access to YouTube’s full content library. Unless YouTube believes you can age backwards, you shouldn’t have to deal with it again.

How unverified YouTube accounts differ from regular YouTube accounts

A lot of “kids” features are harmless. YouTube isn’t allowed to collect information from or show personalized ads to anyone under 13, and I don’t think anyone would object to their ads becoming less creepy. Teen accounts also get “digital wellbeing tools” like break and bedtime reminders. They can be useful for reducing your screentime, although you have to have the self-control to not just click through them and ignore them. (Uh, definitely not speaking from experience, here.)

Unfortunately, there are other features that will seriously impact your YouTube experience. If the algorithm decides you’re a kid or a teen, it’ll restrict what videos you’re allowed to watch. YouTube has guidelines for what it considers adult content, including nudity and gore. The guidelines probably even go too far, banning heavy profanity and lingerie.

YouTubers infamously have to avoid swearing in the first 7–15 seconds of their videos in order to attract the most advertisers, and YouTube sometimes changes its mind on which swear words it considers really bad. It makes sense to hide that stuff from kids 13 and under, but requiring an 18+ verification just for profanity? General audiences can probably handle a little more swearing than YouTube currently allows. (Although YouTube isn’t alone in this kind of censorship; there’s a reason TikTok users have to say things like “unalive” and “sewer slide.”)

But if YouTube’s AI decides you’re under 18, you’ll be barred from even more than just videos involving nudity and profanity. The new system recently prevented one Reddit user from watching The Big Bang Theory, which isn’t exactly rated R. This more likely falls into the category of a 2023 policy change that puts 18+ age restrictions on content portraying “delinquency or negative behaviors, such as cheating on a test, lying for personal benefit, or participating in public pranks.” Also known as the bread and butter of cable sitcoms.

So what if everyone adhered to 100% of YouTube’s content guidelines and avoided anything hypothetically offensive, shocking, or delinquent? YouTubers can be punished if their content is too child-friendly, as well. If a YouTube video gets labelled “for kids,” it gets its comments shut off and loses monetization.

That’s probably why Disney uploaded so many kids’ videos without proper labels, and is now getting fined $10 million over it.

Is YouTube trustworthy?

YouTube is owned by Google, which is part of the fourth-biggest company in the world and as legit as it can get. But that doesn’t mean you should always do everything Google asks of you. Any company can get hacked—like the women’s safety app whose ID verification selfies and user addresses leaked in July. Big-name companies like PayPal have security issues, too. And not even Google is immune.

For example, earlier this year, Google warned users that one of its databases may have been compromised, leading to an increase in phishing attacks impersonating Google support. I begrudgingly went through my accounts and changed all my passwords, although personally, I haven’t seen any emails impersonating Google come through my inbox. At least, not that I can tell. If even Google is vulnerable to cyber attacks, can it really keep user data safe?

And even if Google never gets hacked again, do we know it’ll be 100% responsible with the data—the government IDs, the credit cards, and the selfies— it collects? Most tech companies train their AI without asking permission from original sources. Did you consent to Meta using your Facebook and Instagram posts to train its assistant? AI is a largely unregulated industry right now. What’s stopping Google from feeding your verification selfies into Gemini?

Should you give YouTube your ID?

Personally, I’d feel very reluctant to share my ID with YouTube. I stopped using Venmo in 2021 when they started asking users for their SSNs and IDs. (Sorry to anyone I owe money: You’re getting cash.) And I’d rather take my chances with an entirely new YouTube account than trust a company with information they really shouldn’t need.

Unlike Venmo in 2021, though, YouTube provides users the option to verify their age with a selfie or credit card. I don’t know how I feel about the selfie—it seems much less sensitive than an ID or credit card, but if the AI failed to verify you based on your watch history, who’s to say it’s going to tell your age correctly from an awkward photo? What if you just have a good skincare routine? (If so, share.)

The credit card option is interesting—it works because in the U.S., at least, you need to be 18 to have one. Hypothetically, Google might already have yours on file if you signed up for a YouTube Premium subscription or have ever made a purchase on an Android device through the Google Play store. Honestly, as shady as it sounds, I think I’d be more comfortable confirming my age with a credit card than a government ID. It’s way easier to replace immediately after.

But if your selfie fails, you don’t have a credit card, and you still need to verify your identity, you should keep track of which services have access to something as private as your ID. If you give your government ID to a service like YouTube, Discord, or even Venmo, make note of it somewhere. The next time one of their systems gets hacked, you’ll know if you’re at risk.

And don’t forget about the existence of Google Takeout. It’s a free program that allows you to download any information Google has associated with your account, from Google Drive files to activity logs to YouTube watch history. It won’t delete any information the company has on you, but it’s a great resource for figuring out what they already know. And chances are, they already know a lot.

Your go-to experts for all things TV and streaming

At CableTV.com, we’re passionate about entertainment. If it’s streamable or watchable, we’re interested. That’s why we spend all day researching prices, policies, and press releases—to bring you the best guides the internet has to offer. At least, we’d like to think so. For more on who we are, check out our About Us page. For more on what we do, head over to How We Rank.