How Anyone Can Detect AI Voices and Avoid Being Scammed

Trusting what's real and what's fake on the internet has become harder than ever. With AI everywhere, the lines are blurrier when it comes to determining which voices are real and which are fake. Luckily, certain signs can reveal everything you need to know.

 

6. No natural pauses

In some cases, the lack of natural pauses is a good way to tell if a video is AI-generated — and the same goes for audio. Newer, more expensive models tend to not have this problem, but older AI versions and cheaper models aren't on the same level.

Of course, not having a natural pause can also be the result of poor editing. But in many cases, it's the result of older versions of AI software. While people can still create with these models today, you should especially use this tip for content published in 2023 or earlier.

5. Random errors in voice

AI audio generation has come a long way, and it's easier than ever to create something that accurately resembles real life. But at the same time, AI isn't perfect, so you can often spot errors and other issues if you listen closely enough.

 

If you're listening to multilingual audio, errors become more apparent. For example, the "gracias" in the video below — which was generated using AI — isn't pronounced correctly. In similar videos, you can sometimes notice errors if the speaker's accent isn't their native language.

4. Inconsistency when enabling fast play mode

Playing content faster than it was recorded isn't just optimal for absorbing information faster in some situations. If you're trying to determine whether the audio you're hearing is generated by AI, playing content at 1.25x or faster can make it obvious that you're not hearing a real human voice.

When you play AI-generated audio at a faster than normal speed, it often sounds quite robotic. Sometimes it feels like you're listening to a meme rather than something real.

3. No breathing sound heard

Aside from a lack of natural pauses, another potential sign that you're listening to AI audio—especially if it's a podcast or video—is that you can't hear any breathing. Even if something has been heavily edited, you can still often hear the person breathing if it's their real voice and not an AI-generated one.

Aside from not hearing breathing, you won't hear anything similar when they're actually speaking. For example, you'll rarely hear someone sigh when the voice is AI-generated. The video below, which was developed using AI, is an example.

 

2. Voice is too emotional

While AI voices are getting better at sounding more human than ever before, they aren't necessarily good at expressing the full range of human emotions. One criticism is that AI voices sometimes sound robotic, but the opposite is also true. The short video below, created with AI, is an example.

If you suspect that a voice doesn't sound authentic, it's usually because it was generated using AI.

1. Your intuition says it's AI

Gut feelings evolved to keep humans alive, and while we're often told to ignore them these days, they're often right. While you can run exhaustive analyses and listen to something multiple times to try to decipher whether it's AI-generated or not, sometimes it's worth trusting your gut.

If you're wondering if what you're listening to is AI-generated, you're probably right. You'll become more aware of this the more you listen to AI-generated content, like if you spend a week listening to AI-generated music.

4.5 ★ | 2 Vote

May be interested