AI-generated ads: Türkiye's new rules

Since 1 August 2026, ads aimed at Turkish consumers must clearly disclose AI use that could significantly influence a purchase decision, and any AI-made digital character that cannot be told apart from a human. Using an AI replica of a real person to imply they endorse a product is banned. Targeted ads must explain their criteria, and profiling-based targeting of children is prohibited.

AI-generated ads: Türkiye's new rules

Topic

AI

Published

19 November 2026

Updated

19 November 2026

What changed for AI in Turkish advertising?

Türkiye’s Ministry of Trade made a broad amendment to the Regulation on Commercial Advertising and Unfair Commercial Practices. The amending regulation was published in the Official Gazette (Resmî Gazete) No. 33297 on 1 July 2026 and, under its own commencement article, took effect on 1 August 2026. The same amendment rewrote the rules on consumer reviews; we covered that part in our post on Türkiye’s verified reviews rule. This post covers the provisions on AI, targeting and discounts that affect anyone producing or running ads.

The rules matter to any business advertising to consumers in Türkiye: a Turkish brand, a global brand with a Turkish storefront, or an agency producing campaigns for that market.

Two new provisions deal with AI directly:

  • Article 18, new paragraph 8: where an ad uses AI or other software in a way that significantly affects consumers’ economic behaviour regarding a product or service, or features AI-created digital characters that cannot be distinguished from humans, this must be stated in a clear, understandable and distinguishable way.
  • Article 27, new paragraph 12: an ad may not use an AI-generated digital copy of a real person to suggest, contrary to fact, that the person used, experienced or recommended the product or service.

This is not legal advice. We explain what the regulation says and how marketing teams can build it into their workflows; review your own campaigns with Turkish counsel.

Which AI content needs a disclosure?

The text does not target every AI-touched pixel. It targets two things: AI use that significantly affects the consumer’s decision, and digital characters that pass for human. The table below is a starting point for sorting the most common production scenarios against that test. The answer can change with the ad as a whole and its context.

Scenario Assessment Suggested action
A realistic AI-generated “customer” talks about the product Digital character indistinguishable from a human Add a clear, visible label
An AI-generated product image shows features the real product lacks Can influence economic behaviour Replace it with the real product, or label it; if misleading, drop it
An AI copy of a real celebrity recommends the product Banned under Article 27(12) Do not use
A synthetic human voice speaks as a customer in an audio ad May count as a digital character Disclose it in the audio too
Background clean-up, colour correction, copy drafts May not affect the decision on its own Keep a record; label if borderline
A cartoon mascot that is plainly not human Distinguishable from a human Usually no label needed; check context

The form of the disclosure matters too. The regulation asks for “clear, understandable and distinguishable” disclosure. Small print in the last frame of a video may not meet that bar. Showing the label when the content is shown, in the same language as the ad, is the safer route.

If you also advertise in the EU, the EU AI Act applies on top. Its Article 50 has applied since 2 August 2026 and requires deployers to disclose that deepfake image, audio or video content has been artificially generated or manipulated. The two tests differ, but in practice one labelling policy can cover both markets. For the latest changes to the EU timeline, see our post on the Digital Omnibus delay.

What changed for targeted ads?

New Article 25/A defines targeted advertising as showing ads tailored to specific people or groups by analysing their online behaviour, past preference records, location, demographic data or similar personal data. It applies to sellers, providers and intermediaries acting on their behalf.

There are two rules:

  1. Criteria transparency. Targeted ads are allowed only if consumers get direct, easily accessible information about which criteria were used to show them the ad and how those criteria can be changed.
  2. Children. Where the consumer is known, or reasonably expected, to be a child, profiling based on personal data may not be used for targeted ads.

These rules sit alongside KVKK, Türkiye’s data protection law, rather than replacing it. The legal basis for the data used, the privacy notice and any explicit consent still need their own review, much as they would under GDPR.

Ad platforms’ own “why am I seeing this ad” screens may cover part of the first rule. Personalisation on your own site, app or email campaigns is your responsibility. For flows such as recommendation engines, segment-based pop-ups or basket reminders, design a place that shows which criteria were used and where the user can change them.

How do discount rules interact with AI?

The amendment also changed the reference price for discount claims. Under Article 14, a discount must be measured against the lowest price applied in the ten days before the discount starts. And a price applied in one sales channel may not serve as the basis for a discount in another channel.

This is not an AI rule as such, but it matters for any system that generates campaign copy or price suggestions automatically. A dynamic pricing flow, or one that writes “30% off” headlines, can produce a non-compliant ad if it does not calculate the reference price per channel from the lowest price of the last ten days. During the November and December sales season, running that check automatically before publication is more reliable than trying to catch errors by hand.

The same amendment added a new Article 23/A on social media influencers: influencer ads must include the word “Reklam” (advertisement) or “Tanıtım” (promotion). If you work with an AI-generated virtual influencer, the same post may need both that label and a digital character disclosure.

How can a marketing team put this into practice?

Rather than relying on people to remember each rule, add them to the ad production flow as checkpoints. A workable structure:

  1. Asset register. For every image, video and audio file, record the tool used, whether AI generated it, and who approved it. This lets you show the process if questions come later.
  2. Label rule. Write down which wording is used in which case, where it goes and at what size. For a Turkish audience, for example: “Yapay zeka ile oluşturulmuş görsel” (AI-generated image).
  3. Replica ban. Block AI imitations of real people’s faces or voices entirely unless there is written consent and a legal review.
  4. Targeting card. For each campaign, keep the criteria used and the explanation shown to users in one place. Switch off profiling for audiences that may include children.
  5. Pre-publication checks. Run price-history, label and influencer-wording checks automatically before sign-off; content that fails does not go live.

Some of these checks can be automated with simple rules in the content pipeline. When setting up content automation, keeping the AI draft, human approval and labelling in the same flow is less work than carrying the regulation as a separate checklist. Tying it to a company-wide AI governance policy gives agencies and internal teams the same rules to work from.

Where should you start?

  1. List the ad assets you ran for Turkish consumers in the last three months and mark those made with AI.
  2. Separate those with human-like characters, product images that differ from reality, or imitations of real people.
  3. Put the label wording, placement and format into a written rule, and add it to agency contracts.
  4. Design a criteria explanation for personalisation on your own channels, and switch off profiling where the audience may include children.
  5. Add a step that checks campaign prices against the lowest price of the last ten days, per channel.

The need is sharpest for ecommerce and retail brands with heavy campaign calendars. Corrplus designs AI labelling, approval steps and record-keeping into your ad and content production flow. See our customer experience and digital marketing and AI governance pages for details. To map your own use cases, look at our AI discovery sprint or book a free first call.

Want help with this?

Book a call to apply what “AI-generated ads: Türkiye's new rules” describes to your company. We will listen and agree the first step together.

Frequently asked questions

Do we have to label every AI-generated image in an ad?

The Turkish text targets AI use that could significantly influence consumers' economic behaviour, and digital characters that cannot be distinguished from humans. A cleaned-up background is not the same as an image that shows a product differently from reality. When a case is borderline, labelling is the safer choice; review specific campaigns with Turkish counsel.

Can we use an AI likeness of a celebrity in an ad?

Not to suggest falsely that they used, tried or recommended the product. The amendment bans ads that use an AI-generated digital copy of a real person in that way. Image rights and personal data are separate questions again, so settle them in contract and with counsel.

Are targeted ads now banned in Türkiye?

No. Under new Article 25/A, targeted advertising is allowed if consumers get direct, easily accessible information about the criteria used to show them the ad and how to change those criteria. Profiling based on personal data may not be used to target consumers known, or reasonably expected, to be children.

How do these rules compare with the EU AI Act?

They overlap but do not match. Article 50 of the EU AI Act, applying since 2 August 2026, requires deployers to disclose deepfake image, audio and video content, among other duties. The Turkish rule is an advertising rule focused on consumer decisions. One labelling policy can usually be written to meet both.

Related services

ServiceCustomer experience and marketingWe map the customer journey and bring CRM, marketing automation, search visibility and ad measurement into one plan.
ServiceAI governance and complianceWe inventory your AI use, classify it under KVKK, GDPR and the EU AI Act, and build the rules and technical controls.
ServiceAI consultingWe work out where AI fits in your business, with which data and model, and in what order to build it.

Related articles

Let’s talk about your transformation

Tell us which work repeats every week. We will tell you honestly whether it is worth automating.

How we start

  1. A free first callWe pin down the problem and the goal together.
  2. A 2-week discovery sprintWe look at the data, the process and the tools, and decide where to start.
  3. A written proposalScope, architecture and timeline, in writing.

Example questions

Go straight to

CorrAI answers with AI from the content of this site and can be wrong. Please do not enter personal details. Privacy

Book a free call