Think about you’re scrolling by your TikTok feed. A well-recognized face pops up, somebody you’ve adopted for years. They’re recommending a complement, sounding utterly like themselves, relaxed, real, convincing. You nearly add it to your cart.
Then you definately discover out the video was AI-generated. The creator by no means filmed it. Their voice was cloned. The entire thing was fabricated.
That situation isn’t science fiction. It’s taking place proper now, and Singapore’s regulators have seen. If you happen to’re working model campaigns or working with creators in 2026, right here’s what you really must know.
Singapore Is Getting Severe About AI Content material
For some time, the response to AI-generated content material from most manufacturers was… a shrug. It was a gray space. No person fairly knew the place the traces had been.
That’s modified.
Singapore has moved sooner than most markets within the area. The Infocomm Media Improvement Authority (IMDA) and the Ministry of Communications and Data (MCI) have up to date their joint tips on artificial media and digital promoting, anchored by the On-line Security Act and Singapore’s broader AI governance framework. The Promoting Requirements Authority of Singapore (ASAS) has adopted swimsuit with revised Digital Promoting Tips that apply on to sponsored content material.
What do the principles really say? In plain phrases: if AI was used to generate or meaningfully alter how an actual particular person appears, sounds, or speaks in business content material, it’s important to say so. Prominently. Not in size-6 font on the backside of the caption.
“Materials alteration” covers greater than most entrepreneurs realise. AI-cloned voices? Sure. Face-swapped endorsements? Sure. A magnificence filter that makes a skincare product look simpler than it’s? Additionally sure. The take a look at isn’t whether or not you constructed the entire thing with AI. It’s whether or not AI modified how an actual particular person is perceived in a means that would mislead a viewer.
The “Verified Human” Badge: What It Is and Why It Issues
Platforms haven’t simply been watching from the sidelines. TikTok, Instagram, and YouTube have all rolled out variations of a “Verified Human” badge in Singapore and different regulated markets, and it’s changing into an even bigger deal than most individuals anticipated.
Consider it like a verification checkmark, however for authenticity somewhat than id. When a creator’s content material passes a platform-level scan for AI manipulation, and the creator indicators a declaration confirming their likeness and voice are unaltered, the badge exhibits up on the publish.
For viewers, it’s a fast gut-check. For creators, it’s changing into a sensible requirement. Model contracts in Singapore are more and more together with Verified Human clauses for sponsored posts, as a result of the reputational and authorized dangers of getting this improper are actual sufficient that manufacturers can’t afford to be informal about it anymore.
And right here’s the half that may shock you: content material with the badge is definitely performing higher in sure classes. Finance, well being, parenting, dwelling merchandise. Mainly, any house the place the viewers is already a bit suspicious and is actively in search of causes to belief somebody. Singapore customers are digitally savvy. They discover these items.
Who’s Really Liable When One thing Goes Incorrect
That is the half that tends to make advertising and marketing groups sit up.
Most individuals assume that if a creator produces problematic AI content material, the authorized publicity stops with the creator. It doesn’t.
Underneath Singapore’s Client Safety (Honest Buying and selling) Act, deceptive representations in business contexts can implicate the model that commissioned the content material, the company that accredited it, and the platform that distributed it. Everybody within the chain who knowingly benefited from the content material will be on the hook.
The PDPA provides one other layer. If a creator’s voice or likeness was cloned with out correct consent and knowledge dealing with, there’s a possible Private Knowledge Safety Fee challenge sitting on high of the promoting compliance publicity.
And “the creator agreed to it” received’t prevent. The regulation doesn’t simply require consent between the model and the creator. It requires disclosure to the viewers. These are two utterly various things.
What does a real-world drawback really seem like? It’s not all the time apparent. It might be an AI-cloned model of an area micro-creator’s voice in a product demo they by no means recorded. It might be archival footage of a public determine, synthetically up to date to endorse one thing new. It might be a wellness model utilizing AI to clean out a creator’s pores and skin in a means that inflates what the product can genuinely ship.
None of those really feel like headline-grabbing fraud. They really feel like shortcuts. And in 2026, shortcuts have actual penalties.
Why Belief Is the Actual Problem Beneath All of This
Right here’s one thing that the authorized frameworks don’t fairly seize: individuals are simply drained.
Uninterested in not understanding what’s actual. Uninterested in second-guessing whether or not the particular person recommending one thing really believes in it. Uninterested in discovering out, after the actual fact, that they had been influenced by one thing fabricated.
Singapore’s digital inhabitants sits closely within the 18 to 34 age vary in relation to creator content material consumption, and that’s additionally the group the place belief in social media suggestions has dropped essentially the most sharply over the previous two years. They’re on TikTok, Instagram, and Xiaohongshu always, they usually’ve developed finely tuned instincts for when one thing feels off.
Skeptical audiences don’t convert. They scroll previous.
The creators who’re successful proper now are those leaning onerous into being actual. Actual opinions. Actual environments. Actual reactions, even when these reactions are sophisticated or unflattering to the product. That form of content material builds the type of viewers that manufacturers really need, as a result of the belief is real.
For manufacturers, this implies the advertising and marketing technique must be creator-led, not simply creator-distributed. There’s an actual distinction between briefing a creator to carry out your message and really partnering with somebody whose perspective naturally suits your model. One is a transaction. The opposite is a relationship. And in 2026, solely considered one of them holds up.
What to Do Proper Now
If you happen to’re undecided the place your campaigns stand, right here’s an easy place to begin:
1. Return by your latest sponsored content material
Something that used AI instruments, whether or not for voiceovers, visible enhancement, or generated parts, must be checked in opposition to present IMDA content material requirements and ASAS tips. Higher to search out the gaps your self than have another person discover them for you.
2. Rewrite your creator briefs
Be particular about what AI modifications are and aren’t allowed, what disclosure language is required, and who carries accountability if tips aren’t adopted. Imprecise briefs create obscure accountability.
3. Think twice about who you’re working with
For campaigns in trust-sensitive classes, Verified Human creators aren’t only a compliance checkbox. They’re a real strategic benefit with a Singapore viewers that has seen sufficient to be skeptical.
4. Preserve data
If a criticism goes to CASE, the PDPC, or an promoting physique, the paper path of what was accredited, reviewed, and disclosed turns into your most necessary asset.
Able to Construct One thing Actual?
The manufacturers doing effectively on this surroundings aren’t those scrambling to remain compliant. They’re those who determined that authenticity was value investing in earlier than it grew to become a authorized requirement.
GetKobe helps Singapore manufacturers discover and work with creators who get this, folks whose real voices, and audiences align with what you’re really making an attempt to say. From constructing your first creator partnership to rethinking an present programme for the present panorama, the workforce will help you progress sooner and smarter with out reducing corners on the issues that matter.
In a post-truth feed, essentially the most highly effective factor you possibly can supply your viewers is one thing actual.
CONTACT US
