TikTok's AI Ad Problem Explained: How Brands Generate AI Ads (And How You Can Too)

I've been staring at TikTok ads trying to figure out if they're AI-generated. Some of them are obvious. Others? Even with trained eyes, I can't tell.

The Verge investigated and found something irking: Samsung ran AI-generated ads for the Galaxy S26 Ultra on TikTok without the required AI disclosure labels— even though the same campaign on YouTube had proper disclosures. TikTok's own policy requires AI labeling. Samsung is a member of the Content Authenticity Initiative. Neither of them followed through.

The uncomfortable truth: TikTok can't reliably detect AI-generated ads, and brands know it.

The Problem: TikTok's AI Labeling Is Broken

Here's what happened with Samsung:

• Samsung created AI-generated promotional videos for the Galaxy S26 Ultra
• On YouTube, the videos had proper AI disclosure labels in the description
• On TikTok, the same campaign ran with no AI disclosure
• TikTok's automated detection didn't catch it

Both Samsung and TikTok are members of the Content Authenticity Initiative, which promotes C2PA standards for AI content labeling. Both failed to apply those standards in practice.

How AI Ads Are Actually Made

The technology behind AI-generated ads isn't mysterious — it's just image and video generation models accessed via API. Here's the stack:

This is why Samsung and every other major brand is using AI for ad creatives. The economics are undeniable.

Tutorial: Build Your Own AI Ad Creative Generator

Here's the actual code. This is what brands like Samsung are using (or something very close to it):

Python

# Install: pip install nexaapi
from nexaapi import NexaAPI

client = NexaAPI(api_key="YOUR_API_KEY")

# Generate an AI ad creative for TikTok
response = client.images.generate(
    model="flux",  # or stable-diffusion-xl, etc.
    prompt="A vibrant, eye-catching product advertisement for a smartphone, TikTok vertical format, bold colors, modern lifestyle aesthetic, photorealistic",
    width=1080,
    height=1920,  # TikTok vertical 9:16 ratio
    num_images=4
)

for i, image in enumerate(response.images):
    image.save(f"tiktok_ad_creative_{i+1}.png")
    print(f"Ad creative {i+1} saved — cost: $0.003")

print(f"Total cost for 4 ad creatives: {4 * 0.003:.3f{'}'}")

JavaScript

// Install: npm install nexaapi
import NexaAPI from 'nexaapi';

const client = new NexaAPI({ apiKey: 'YOUR_API_KEY' });

async function generateTikTokAdCreative() {
  const response = await client.images.generate({
    model: 'flux',
    prompt: 'A vibrant, eye-catching product advertisement for a smartphone, TikTok vertical format, bold colors, modern lifestyle aesthetic, photorealistic',
    width: 1080,
    height: 1920, // TikTok vertical 9:16 ratio
    num_images: 4
  });

  response.images.forEach((image, index) => {
    image.save(`tiktok_ad_creative_${index + 1}.png`);
    console.log(`Ad creative ${index + 1} generated — cost: $0.003`);
  });

  console.log(`Total cost for 4 ad creatives: $${(4 * 0.003).toFixed(3)}`);
}

generateTikTokAdCreative();

The Pricing Reality

What This Means for Developers and Marketers

The TikTok/Samsung story isn't just about disclosure policy. It's a signal:

1. AI ad generation is mainstream — If Samsung is doing it, everyone is doing it
2. The tools are accessible — Any developer can build this pipeline today
3. Disclosure is coming — Regulation will eventually catch up; build disclosure into your workflow now