What truly irks me about this situation is the undeniable fact that someone within the advertising chain—be it the brand creating the ad or the platform hosting it—possesses concrete knowledge of whether the content is AI-generated. Yet, that crucial information is consistently withheld from the vast majority of users. This deliberate or accidental obfuscation directly undermines the very principles of transparency that companies claiming to champion AI-labeling initiatives supposedly uphold. If these corporate giants genuinely aspire for such initiatives to succeed and foster a trustworthy digital environment, immediate and decisive action is imperative.
Consider the case of Samsung, a global technology leader. After extensively utilizing AI-generated videos across its social media channels, I began noticing advertisements teasing the Galaxy S26 Ultra’s privacy display feature appearing on my TikTok feed. Intrigued, I discovered that videos from what appeared to be the identical promotional campaign had been published on YouTube, where their collapsed descriptions explicitly disclosed the use of AI tools in their creation. In stark contrast, the corresponding TikTok ads offered absolutely no indication of AI involvement. Furthermore, even regular, non-promoted videos on Samsung’s TikTok accounts frequently lacked AI disclosures, despite those very same videos being clearly labeled as AI-generated on YouTube. This glaring inconsistency points to a fundamental breakdown in either Samsung’s internal content management or TikTok’s enforcement mechanisms.
It is particularly significant to highlight that both Samsung and TikTok are active members of the Content Authenticity Initiative (CAI), a collaborative group dedicated to making content authenticity and transparency “scalable and accessible.” Their stated mission is to achieve this through the industry-wide adoption of the C2PA (Coalition for Content Provenance and Authenticity) standard, which aims to provide verifiable metadata about a piece of content’s origin and modifications. This membership strongly implies that TikTok and Samsung supposedly share common ideals regarding the transparent labeling of AI-generated content. Therefore, if Samsung knowingly employed AI to produce its promotional videos, it bore the responsibility to inform TikTok during the ad submission process. Conversely, if TikTok was indeed informed, it had a clear obligation, under its own advertising policies, to ensure its users were made aware of the AI origin. The failure at either or both of these points represents a serious breach of trust and policy.
TikTok’s advertising policies explicitly permit advertisers to use content “significantly” edited or generated by AI, provided they disclose this fact. This disclosure can be achieved either by applying TikTok’s native AI label or by incorporating a disclaimer, caption, watermark, or sticker of the advertiser’s choice. The policy meticulously defines “significantly modified by AI” to encompass content altered beyond minor tweaks or enhancements. This includes:
- Content that contains images, video, or audio that are completely AI-generated.
- Showing the primary subject performing an action they didn’t actually do, such as dancing.
- Making the primary subject utter words they didn’t actually say, using AI voice-cloning technology.
Given these clear guidelines, the central question remains: what precisely went wrong in the Samsung instance? My attempts to obtain clarification from Samsung were met with silence. TikTok, when contacted, directed me to its general AI labeling requirements for advertisers and its C2PA partnership, but pointedly declined to provide an on-record statement explaining why Samsung’s AI-generated ads seemingly bypassed these disclosure requirements. This lack of accountability from both parties leaves me, and by extension the public, completely in the dark regarding the precise failure point in this critical transparency process.
However, a recent development offered a glimmer of hope, albeit a reactive one. Earlier this week, I observed that TikTok ads promoted by Cazoo, a UK-based used car retailer, which I had previously encountered without any disclosure, now featured a message reading “advertiser labeled as AI-generated” at the bottom, adjacent to the “Ad” identifier. My initial suspicion that these ads were AI-generated stemmed from a series of bizarre visual distortions—such as a dentist’s drill morphing into different shapes and inexplicably jumping between hands—that lacked any rational editing explanation. The belated application of a label, while welcome, underscores the reactive nature of the current system; it shouldn’t require individual scrutiny and flagging to enforce basic transparency.
It remains unclear if Samsung’s ads on TikTok have undergone a similar retroactive update, as they have not appeared in my feeds for several days. Nevertheless, the overall state of AI transparency across Samsung’s various TikTok accounts is undeniably chaotic. Some videos display TikTok’s official AI label, others include a manual disclosure embedded within the video’s fine print, while numerous demonstrably AI-generated examples carry no disclosure whatsoever. This inconsistency suggests a haphazard approach rather than a robust, systematically enforced policy.
The broader challenge of identifying AI-generated content at scale is formidable. There is currently no perfectly reliable technological solution for definitively distinguishing AI-made content from human-made content. I have extensively critiqued the inherent flaws in authentication standards like C2PA Content Credentials, SynthID, and other provenance-based systems. Their efficacy hinges on universal adoption, a scenario that remains stubbornly out of reach in the sprawling and often unregulated digital landscape. This deficiency in reliable identification technology is particularly problematic in our current geopolitical climate, where the struggle to differentiate factual information from synthetic fabrication has profound implications for public discourse and democratic processes.
However, this argument primarily applies to online content in general. Advertising, in contrast, operates within a heavily regulated industry that is fundamentally expected to adhere to a distinct set of rules. Many of these regulations were specifically established to safeguard consumers from being misled or outright deceived by advertisers. Examples abound, from laws preventing cosmetics companies from using false eyelashes to sell mascara, to strict guidelines on health claims for food products. TikTok beauty influencers, such as Mikayla Nogueira, have learned the hard way that these rules extend to their promotional activities, and that audiences react strongly to dishonest shilling tactics, as seen in the backlash over her mascara reviews.
While AI-generated videos are not inherently misleading, their potential for deception is significant. Concerns regarding advertising transparency have already spurred legislative action in various jurisdictions. The European Union, China, and South Korea, for instance, have introduced or are actively developing mandatory labeling requirements for AI content in promotional materials. Even companies that have not formally pledged to support AI transparency initiatives risk substantial fines and reputational damage if they fail to adapt and implement robust disclosure practices. The global trend is clear: greater transparency in AI-driven advertising is becoming a legal and ethical imperative.
If major online platforms like TikTok and prominent advertisers such as Samsung cannot be relied upon to be forthright about AI usage within such a regulated environment, the integrity of the entire digital advertising ecosystem is jeopardized. It creates a dangerous precedent where any entity could potentially advertise any form of nonsense without genuine accountability. While I am heartened that some ad-specific AI labels are now appearing on TikTok—particularly after I directly flagged specific ads to the companies involved—this should not be the modus operandi. This is a fundamental two-way system that ought to be robustly implemented and rigorously enforced by default, without requiring individuals like myself to meticulously scrutinize every single advertisement that appears in our feeds. The onus for transparency and accountability must lie squarely with the platforms and the advertisers, ensuring a digital space where trust is built, not eroded.
Post Views: 2
