What Changed
On July 1, 2026, Turkey's Ministry of Trade published amendments to the Regulation on Commercial Advertising and Unfair Commercial Practices in the Official Gazette (Issue 33297), taking effect August 1, 2026 after a one-month transition window. The amendment does three things at once. It elevates informal influencer-marketing guidance into binding law: paid social media posts must now carry a visible "Reklam" (Advertisement) or "Tanıtım" (Promotion) label plus the advertiser's identity, in a size and color that doesn't require scrolling to see. It creates Turkey's first explicit AI-in-advertising disclosure rule, requiring advertisers to flag when a "digital character" or AI-generated likeness appears in a way that could shape a consumer's economic decision, and it bans using an AI replica of a real person to imply they used or endorsed a product. And it gives "targeted advertising" — ads built from behavioral, location, or demographic profiling — a legal definition for the first time, requiring platforms to disclose why a user was shown an ad, what criteria drove it, and how to change those criteria. Profiling-based targeting of children is banned outright. All of it sits under the enforcement authority of the Advertising Board (Reklam Kurulu), which can order an ad pulled, corrected, or fined (Ministry of Trade).
The Case for It
The strongest argument for this rule isn't hypothetical. Turkey's influencer economy has operated for years on an ambiguous standard — a non-binding Ministry guideline that platforms and creators could and did ignore, because ignoring it carried no real cost. Undisclosed sponsorship is a genuine consumer-protection problem: a viewer who can't tell a product recommendation from a paid placement is being deceived regardless of intent, and that harm scales with follower count. The AI-disclosure clause closes a gap that didn't exist when the base regulation was written — synthetic testimonials and AI "spokesmodels" are now cheap enough to produce that a bright-line rule against passing one off as a real endorsement is a reasonable, narrowly-targeted response, not overreach. And the child-profiling ban is close to uncontroversial: it tracks the direction most serious regulators (the EU's DSA, UK's Age Appropriate Design Code) have already moved, and it targets a specific, defined harm — profiling minors for ad delivery — rather than banning a broad category of speech.
Where the Regulation Overreaches
The trouble is what sits alongside those defensible provisions. "Targeted advertising" is now a defined legal category enforced by a government board with the power to suspend an ad and fine the publisher, and the definition — any ad shaped by "online behaviour, previous preferences, location information, demographic data or similar personal information" — covers most of programmatic advertising as it currently exists, not just the aggressive profiling cases regulators usually mean to target (Pi Legal Consultancy). That's a wide net administered by an executive-branch board, not a court or an independent regulator insulated from political pressure the way KVKK (Turkey's data protection authority) or a judicial process would be. The Advertising Board isn't a hypothetical enforcer, either: in the first eight months of 2026 it reviewed roughly 26,000 applications and imposed 218,478,014 TL in fines for deceptive advertising and unfair commercial practices, including a single August 13 session that fined 87 companies 32,683,631 TL (Ministry of Trade enforcement notice). Social media and internet advertising carry a fine ceiling of roughly 10,837,065 TL per violation under the Board's 2026 penalty schedule — among the highest per-post exposure categories in the entire framework, well above local TV or print (penalty schedule analysis). For an individual creator or small brand, that's not a deterrent calibrated to the offense — it's an existential threat calibrated to force self-censorship well beyond what disclosure law requires, especially given the ambiguity in what counts as "targeted."
The Real Risk Is the Discretion, Not the Disclosure
Mandatory sponsorship labels and a ban on deceptive AI endorsements are proportionate, defensible rules that plenty of Western jurisdictions already enforce in some form. What should concern anyone tracking Turkey's broader approach to online speech is the combination: a vague, maximalist definition of "targeted advertising," fine caps steep enough to chill ordinary ad-tech practice, and an executive board — not a court — deciding case by case what counts as a violation. Turkey's Advertising Board already functions as an active enforcer, not a paper tiger; folding an ambiguously-defined category of everyday digital advertising into that same fast-moving, high-penalty system invites the kind of discretionary pressure on platforms and creators that goes well past protecting consumers from deception. Ankara didn't need to choose between the two: a narrower definition of targeting — one that actually distinguishes ordinary contextual ad delivery from the profiling practices worth restricting — would have delivered the same consumer protection without the same chilling radius. Regulators elsewhere weighing similar rules on influencer and AI disclosure should take the disclosure mandate and leave the vague catch-all behind.