Why can I spot AI-generated ads on TikTok when it can’t?

I’ve been trying to figure out if the ads popping up in my TikTok feed are crafted using generative AI tools. As someone who spends a lot of time analyzing images and videos for the usual signs of synthetic creation, some of the promotions I’ve encountered have definitely raised my eyebrows. For weeks, I didn’t spot any ads with the required AI disclosure mandated by TikTok’s advertising guidelines, leaving me in the dark.

What frustrates me is that someone out there knows whether the content is AI-generated, but they’re not sharing that info with the rest of us. If companies that advocate for AI labeling genuinely want these initiatives to succeed, they need to step up.

Take Samsung, for instance. After flooding its social media with AI-generated videos, I started seeing ads for the Galaxy S26 Ultra’s privacy display feature on TikTok. Videos from what seems to be the same campaign were uploaded to YouTube, complete with disclosures in their descriptions indicating that AI tools were used. In contrast, the TikTok ads provided no hints about AI involvement. Even regular videos on Samsung’s TikTok account, which aren’t promoted as ads, lack AI disclosures, despite being labeled as AI-generated on YouTube.

It’s worth noting that both Samsung and TikTok are part of the Content Authenticity Initiative, which aims to enhance content authenticity and transparency by promoting the industry-wide adoption of C2PA. This suggests that TikTok and Samsung should ideally align on the labeling of AI content. If Samsung knowingly used AI for its videos, it should have informed TikTok when submitting the ads. If TikTok was made aware, it should have ensured users were informed, in line with its own advertising policies.

Advertisers on TikTok are only allowed to use content that is “significantly” edited or generated by AI if they disclose that fact. This can be done by applying TikTok’s AI label or adding a disclaimer, caption, watermark, or sticker of their choice, according to the platform’s advertising policy. They define “significantly modified by AI” as content altered beyond minor tweaks, including:

– Content that features completely AI-generated images, videos, or audio.
– Depicting the main subject doing something they didn’t actually do, like dancing.
– Making the main subject say something they didn’t actually say, using AI voice-cloning.

So, what’s the deal?

Samsung didn’t respond to my inquiries. TikTok directed me to its AI labeling requirements for advertisers and its C2PA partnership but didn’t provide a statement on why Samsung’s AI-generated ads slipped through without a disclosure. I’m still unclear on where the transparency process broke down.

Earlier this week, I noticed a new twist — TikTok ads from a UK-based used car retailer that I had previously seen without a disclosure now feature a message stating “advertiser labeled as AI-generated” at the bottom, next to the “Ad” tag. I had already suspected these ads were likely AI-generated due to their odd visual distortions, like a dentist’s drill morphing into different shapes and switching hands.

I can’t tell if Samsung’s TikTok ads have received a similar update since it’s been a few days since any were promoted to my feed. The AI transparency across Samsung’s TikTok accounts is generally inconsistent — some videos have TikTok’s AI label, others include a manual disclosure in the fine print, and several AI-generated examples lack any disclosure whatsoever.

Currently, there’s no reliable technological solution for identifying AI-generated content or even human-made content at scale. I’ve spent a lot of time discussing the shortcomings of authentication standards like C2PA Content Credentials, SynthID, and other provenance-based systems that aim to inform users about how content was created. These systems require widespread cooperation to function effectively, which simply isn’t happening. This is a significant issue when people are trying to discern what’s real in today’s complex geopolitical landscape.

However, this issue extends beyond general online content; advertising is a regulated industry that should adhere to a different set of rules. Many of these regulations exist to protect consumers from being misled or deceived by advertisers, such as laws preventing cosmetics companies from using false imagery to sell their products. TikTok beauty influencers have learned the hard way that these rules apply to them, as audiences often react negatively to dishonest promotion tactics.

While not all generated videos are misleading, concerns about advertising transparency have led the EU, China, and South Korea to implement labeling requirements for AI in promotional materials. Companies that don’t commit to AI transparency initiatives could face fines if they don’t get their act together.

If major platforms like TikTok and advertisers like Samsung can’t be upfront about AI usage in such a regulated environment, it opens the door for anyone to promote whatever they want without accountability. I’m glad to see that some AI labels are starting to appear on TikTok ads after I flagged them to the companies involved. However, this should be a straightforward two-way system that is already effectively implemented and enforced, without needing vigilant users to scrutinize every ad in their feeds.