I usually record these episodes about a week ahead of time. Not this one. I recorded this on a Monday because of everything that happened over the weekend, and honestly, there was so much misinformation flying around that it couldn’t wait until my normal schedule.
So let’s break down what’s actually going on.
What Is the “AI Info” Label?
Instagram has started labeling content as “AI info.” You’ll see it in the same spot where the song title, original audio, or location tag usually shows up. It actually cycles between all of those, the music, the location, and now AI info, as you’re watching a reel or swiping a carousel.
And I’m genuinely glad Instagram is doing this. We now have entire creator accounts with millions of followers that aren’t real people. Most of us who live on this platform can spot an AI-generated person pretty easily. A lot of our family and friends can’t. That gap is only going to get harder to close as AI gets better, so a labeling system is a good thing.
Why Are People Getting Flagged for Barely Using AI?
Here’s where it gets messy. One creator this weekend said all of her content was getting flagged, and the only AI tool she’d used anywhere was Canva’s background remover. Canva has actually come out and said that specific tool isn’t supposed to trigger the AI info label. But she’s not the only one this happened to.
Closed captions are another likely culprit. If you’re using AI to auto-generate your captions (and a lot of us are), that could be enough to trigger the flag too.
My honest take is that this is brand new, it’s in a testing phase, and it’s probably overcorrecting right now. As Instagram’s own systems get better at telling the difference, I expect this to get less sensitive over time.
The New Text Watermark Thing
This is the part that actually matters most for how you create content going forward. Claude, ChatGPT, Gemini, and other AI tools started implementing hidden watermarks in their text output this past week.
Here’s what that means practically. If you copy and paste something directly from one of these tools into your caption, that watermark travels with it. Some creators are also saying the watermark can show up even if you type the words out yourself instead of copy-pasting, though that part is less confirmed.
My recommendation hasn’t really changed because of this, it’s just more important now. Use AI to generate ideas or a structure, sure. But don’t copy and paste it straight into your caption. Take the idea, take the bones of it, and put it in your own words, the way you’d actually text a friend. Yes, it takes a little more time. But it also just makes your content sound more like you, regardless of the AI question.
Does Getting Flagged Actually Hurt Your Reach?
No. Meta has directly confirmed that being labeled AI info does not affect how many people see your content. It’s still shown to the same number of people. What happens after that, whether someone chooses to engage with something tagged as possible AI content, is up to them.
I don’t think this is going to be a huge problem long term, honestly, because at this point almost everyone is using AI in some form.
What I’d Actually Do If You’re Worried
Go back to basics. Talk to the camera. Use real photos and videos of yourself. Use your own words.
And here’s an interesting wrinkle. Even fully human-written paragraphs can sometimes get flagged, because AI models were trained on human writing in the first place, including the way a lot of us naturally talk and write. I’ve personally had to stop using em dashes in my own content and my clients’ content, not because I started using them for AI, I’ve used them for years, but because I didn’t want anyone assuming I did.
It’s frustrating. But we’re all figuring this out together right now.
Recap
- Instagram’s new AI info label shows up where the audio or location tag usually is
- Small tools like background removers or auto-captions might be triggering false flags right now, and that should improve over time
- AI text tools now embed hidden watermarks, so don’t copy and paste directly into your captions, rewrite in your own voice instead
- Getting labeled AI info does not reduce your reach, Meta has confirmed this directly
- If you’re worried, go back to basics: real footage, real photos, your own words
Your Action Challenge
This week, get on camera for a face-talking story or reel. Think of a question people ask you all the time, put your phone down, and just start talking about it. Talk for 10 minutes if you need to. Then trim it down to a minute and a half or two minutes and post it as a reel.
If you want to use AI here, this is a great spot for it: feed your transcript into whatever AI tool you use and ask it to flag the strongest 90 seconds. Then trim it yourself. I’d bet that piece of content does better than you’re expecting.
Links for blogs
Resources:
The Plug & Post Instagram Content Membership: Instagram content ideas, picked by me (an IG strategist) every week based on what’s working RIGHT NOW… so all you have to do is film and press post! – Join here!
100 IG stories that sell – $17
Say Hi on instagram! @nicoleialeggio
Try Manychat- 1 month free offer HERE
