A nyone else getting excessive “AI Content Detected” messages on YouTube videos - I’m getting repeated pop up warnings from videos I KNOW aren’t AI, and when I dismiss the pop up it generally comes back after a couple of minutes.
I think that the AI/Deepfake detection has been turned a tad too aggressive. I’m almost at the point of disabling it (and a protection feature that drives users to deliberately disable it isn’t exactly good).
Deep Fake protection is in its very early stages of production. I would expect a great amount of false positives and negatives for some time.
If it were me I would turn it off unless I wanted to do some beta testing for Norton.
I’ve not yet run into any of this AI Content Detected messages, yet. Is there any where that you can find in the message that allows you to challenge, or report the detection to help ‘teach’ the AI?
Yes, there is a way to report false positives, but the system seems to ignore them.
Bonjour,
Etant établit en France, je n’ai pas encore rencontré ce type de message. Par ailleurs, je ne sais pas si cette fonctionnalité est active pour les clients français.
Do you report them to YouTube or Norton? Either way, any reports are not likely to be addressed in real time.
Reported to Norton, via the App itself.
Thanks for the clarification.
Bonjour,
Je reviens vers vous pour dire qu’aujourdhui, j’ai découvert en debut d’après-midi que désormais, la protection contre les escroqueries pro (les SMS sécurisés ; les appels sécurisés et les emails sécurisés) était incluse dans mon abonnement Norton 360 Advanced. Je suis contente de cette avancée même si je trouve un peu dommage que le nombre d’emails protégés soit limité à 5, alors que les concurrents que sont McAfee et Bitdefender proposent la protection de 10 emails dans leurs abonnements premium respectifs.
Par ailleurs, je dois préciser que j’ai dû désinstaller puis réinstaller Norton 360 sur mon smartphone pour pouvoir bénéficier de la protection contre les escroqueries Pro.
Je n’ai pas encore reçu de message d’alerte au sujet de YouTube.
With the latest update of Norton 360, we have noticed that deepfake protection has started to generate excessive detection on AI generated content. While our team is looking at tuning the AI model, you can turn off the setting “Send me AI content notifications” to avoid excessive notifications. We will update the thread once we have a solution in place.
Thanks
Hello @Selvakumar.S
https://community.norton.com/t/norton-26-2-keeps-crashing/513602
https://community.norton.com/t/norton-26-2-keeps-crashing/513602/35
Norton’s AI/deepfake warning behavior on YouTube feels like a textbook example of security theater drifting into product malpractice.
We understand the desire to protect people from scams. Deepfakes are of course real. Fraud is real. Malicious ads, cloned channels, fake investment pitches, and social-engineering videos are absolutely a problem.
But interrupting ordinary YouTube viewing with alarming popups because a video discusses AI, agents, payments, or uses audio that a detector does not like is not protection. It is noise.
A warning that says “Deepfake scam detected” carries serious weight. It implies the content itself is fraudulent. If the system is merely guessing based on audio patterns, topic keywords, or risk signals, then the language needs to reflect that uncertainty. Otherwise, Norton is effectively accusing legitimate creators of running scams based on an opaque classifier the user cannot meaningfully evaluate.
That is not good security. Good security helps users make better decisions. Bad security trains users to ignore warnings.
A better approach would be:
“Potential AI-generated or manipulated audio detected. Be cautious with links, payments, downloads, or financial claims.”
That would be useful. It would warn without overclaiming. It would protect users without smearing creators. It would encourage critical thinking instead of panic-click dismissal.
Right now, this feels less like consumer protection and more like ambulance-chasing branding: take a real fear, wrap it in a dramatic alert, and use it to make the product look indispensable.
Norton should absolutely help users avoid scams. But if it wants to flag content, it needs to be precise, transparent, and clear about what its AI actually detected.
The irony is hard to miss: Norton is warning users not to trust AI-generated content while asking users to blindly trust Norton’s own AI-generated accusation.
