Artificial intelligence can now create an entire advertising campaign in seconds.

A company can ask a generative AI system to write product claims, produce photographs, create an artificial spokesperson, generate customer testimonials and even produce a video showing the product being used.

But imagine that the AI-generated advertisement says:

“Clinically proven to reduce hair loss by 90%.”

No such clinical study exists.

Or an AI-generated virtual customer says:

“I lost 12 kilograms in one month using this product.”

The person does not exist.

Or a famous actor is digitally recreated by artificial intelligence and appears to say:

“I use this investment platform myself.”

The actor never said or did anything of the kind.

Who is legally responsible?

Can the brand defend itself by saying:

“ChatGPT generated the text. We did not know it was false.”

Under Turkish consumer and advertising law, this will generally be a very weak defence.

The central principle is straightforward:

Artificial intelligence is a production tool. Using AI does not normally transfer the advertiser’s legal responsibility to the software.

This issue has become particularly important following Türkiye’s major amendment to the Commercial Advertising and Unfair Commercial Practices Regulation (Ticari Reklam ve Haksız Ticari Uygulamalar Yönetmeliği), published on 1 July 2026 and effective from 1 August 2026.

The amended Regulation now contains express rules concerning AI-generated advertising. Where AI or another software is used in advertising in a way that may significantly affect consumers’ economic behaviour, or where AI-generated digital characters indistinguishable from humans are used, that fact must be disclosed clearly and distinguishably. The Regulation also expressly prohibits advertisements in which an AI-generated digital copy of a real person falsely gives the impression that the person personally used, experienced or recommended a product or service.

The consequence for businesses is significant:

Türkiye has moved from applying general misleading-advertising principles to AI content to expressly regulating AI-generated advertisements themselves.

1. AI-Generated Advertising Is Still “Advertising”

The first misconception is that content automatically created by software is somehow outside traditional advertising law.

It is not.

Article 61 of Consumer Protection Law No. 6502 requires commercial advertising to comply with legal standards and prohibits misleading advertisements and advertisements exploiting consumers’ lack of knowledge or experience.

More importantly, advertisers are required to prove the accuracy of claims appearing in their advertisements, and advertisers, advertising agencies and media organisations are all required to comply with Article 61.

The Commercial Advertising Regulation likewise applies broadly to consumer-directed commercial advertising and unfair commercial practices.

Therefore, it makes little difference whether the sentence:

“Our product eliminates 95% of wrinkles”

was written by:

• a marketing employee;
• an advertising agency;
• an influencer;
• ChatGPT;
• another generative AI system;
• or an automated advertising algorithm.

Once the business approves and publishes the claim as part of its commercial communication, the claim must comply with advertising law.

The legally relevant question is not:

Who typed the sentence?

It is:

Who used that sentence commercially to influence consumers?

2. The Brand or Advertiser Will Usually Be the First Party in the Liability Chain

In most cases, the advertiser is the most important responsible party.

Suppose a cosmetics company instructs an AI system:

“Write a convincing Instagram advertisement for our anti-ageing cream.”

The AI responds:

“Dermatologists confirm that this cream reverses skin ageing within seven days.”

No such evidence exists.

The company publishes the statement without verification.

The advertiser cannot ordinarily avoid responsibility by arguing that the AI invented the claim.

Under Article 61/6 of Law No. 6502, the advertiser bears the burden of proving the accuracy of claims made in its commercial advertisements.

This produces one of the most important compliance principles for generative AI advertising:

A business must verify AI-generated factual claims before publication.

Generative AI systems may generate plausible but incorrect statements.

From an advertising-law perspective, this technical phenomenon does not convert an inaccurate commercial claim into a lawful one.

3. “The AI Hallucinated” Is Not a Consumer-Law Defence

Assume an AI system produces the following statement:

“According to Oxford University research, our supplement increases concentration by 47%.”

The marketing department does not check the alleged study.

The claim is published.

Later, it becomes clear that the study never existed.

The company says:

“The AI hallucinated the source.”

That explanation may describe how the false information appeared.

It does not necessarily answer who is legally responsible for publishing it.

The advertiser deliberately placed the claim before consumers.

The advertiser also carries the statutory burden of substantiating advertising claims.

Accordingly, AI-generated claims should be treated internally in the same way as claims submitted by an external advertising agency:

they must be legally and factually reviewed before publication.

4. Türkiye Now Has a Specific AI-Advertising Disclosure Rule

The 2026 amendment is particularly important.

Article 18/8 of the Commercial Advertising Regulation now provides that where AI or another software is used in advertising in a way that significantly affects consumers’ economic behaviour concerning a product or service, or where AI technologies are used to create digital characters indistinguishable from humans, the AI/software use must be disclosed clearly, understandably and distinguishably.

This rule means that businesses should not simply ask:

“Is the content false?”

They must also ask:

“Does the consumer need to be told that AI was used?”

For example, an advertisement containing an obviously animated cartoon character may raise a different issue from an advertisement showing an entirely synthetic but photorealistic “doctor” who appears to be a real human professional.

Where consumers could reasonably perceive an artificial character as a genuine individual, transparency becomes central.

5. Deepfake Testimonials Are Now Expressly Restricted

The 2026 Regulation goes further than simply requiring AI labels.

The Regulation expressly prohibits advertisements in which the AI-generated digital copy of a real person falsely gives the impression that the person:

• personally experienced a product;
• personally used a product; or
• recommends the product or service.

Consider this example.

An online investment company generates an AI video using the digital likeness and voice of a well-known economist.

The video says:

“I personally invest through Platform X. It is the safest investment platform in Türkiye.”

The economist never used the platform and never made the statement.

This is substantially more serious than simply failing to add a small “AI-generated” notice.

The underlying testimonial itself falls within the specific prohibition.

A disclosure such as:

“AI-generated video”

would therefore not necessarily cure the unlawful content.

6. Disclosure Does Not Make a False Claim Lawful

This distinction is essential.

Suppose an advertisement states:

“AI-generated advertisement. This medicine cures diabetes.”

If the underlying health claim is legally prohibited or scientifically unsupported, informing consumers that AI created the advertisement does not make the claim lawful.

AI transparency and advertising truthfulness are separate obligations.

An advertisement may therefore violate Turkish law because:

1. AI involvement was not disclosed where disclosure was required;
2. the underlying product claim was misleading;
3. the scientific claim could not be substantiated;
4. an artificial testimonial was presented as genuine;
5. the advertisement concealed material information;
6. or several of these violations occurred simultaneously.

The Regulation provides that information capable of affecting purchasing decisions must be presented in a form and within a time that the average consumer can perceive, and the essential components of the principal advertising promise must be communicated clearly.

7. Can an Advertising Agency Also Be Responsible?

Potentially, yes.

Article 61/7 of Law No. 6502 expressly provides that:

• advertisers;
• advertising agencies; and
• media organisations

must comply with the commercial advertising requirements.

Suppose a company hires an advertising agency.

The agency uses AI to create a campaign stating:

“Independent tests prove that this cleaning product removes 100% of bacteria.”

Neither the agency nor the client checks whether such tests exist.

The advertisement is published nationally.

The brand should not assume:

“The agency created it, so the agency is responsible.”

Likewise, the agency should not assume:

“The client approved it, so only the client is responsible.”

Advertising law may expose multiple participants in the advertising chain.

Contracts between the company and agency can allocate internal responsibilities, indemnification obligations and approval procedures.

But an internal contract does not necessarily eliminate statutory responsibility toward regulators or consumers.

8. What If the Agency Says “Our AI Tool Generated the Claim”?

That defence faces the same fundamental problem.

Professional advertising agencies are expected to operate with appropriate professional care.

If an agency uses generative AI to create factual advertising claims, the agency should establish a review system.

High-risk claims include statements such as:

• “clinically proven”;
• “doctor recommended”;
• “the safest”;
• “number one in Türkiye”;
• “guaranteed return”;
• “zero risk”;
• “scientifically demonstrated”;
• “100% effective”;
• “government approved”;
• “certified”;
• and “used by millions.”

These are not merely creative expressions.

They are factual propositions capable of requiring evidence.

9. The Advertising Board Has Already Rejected Advertising Based on Unreliable AI Statements

Even before the 2026 express AI rules, the Turkish Advertising Board had dealt with advertising claims based on ChatGPT outputs.

In one decision, a company relied on ChatGPT-related statements to create an impression of superiority over competitors. The Board observed that AI-generated answers did not have the necessary certainty or reliability to substantiate objective superiority claims and concluded that the advertisements were misleading where they were not supported by objective research evidence.

This decision illustrates a critical principle:

“AI says we are the best” is not evidence that you are the best.

An advertiser cannot outsource the evidentiary burden to a language model.

If the advertisement contains an objectively measurable claim, the advertiser should possess appropriate supporting evidence.

10. What If an Influencer Publishes the AI-Generated Advertisement?

The liability analysis becomes more complicated.

Suppose a brand provides an influencer with an AI-generated script:

“This collagen supplement completely eliminates joint pain.”

The influencer reads it in a TikTok video.

Several legal questions arise.

Was the claim false?

Was the commercial relationship disclosed?

Did the influencer present personal experience that never occurred?

Was the brand involved?

Was AI used to create a synthetic voice, image or testimonial?

Türkiye’s amended Regulation now expressly regulates advertising through social-media influencers.

Such advertisements must be clear, understandable and distinguishable as advertising, and where benefits such as payment, free or discounted products or services are provided, the advertising character must be disclosed. The Regulation requires prescribed “Reklam” or “Tanıtım” disclosures and sets visibility requirements for those disclosures.

Accordingly, both the underlying product claim and the manner in which the commercial relationship is presented must be examined.

11. An Influencer Cannot Pretend to Have Used a Product Merely Because AI Wrote the Script

Consider a more serious example.

A cosmetic clinic pays an influencer.

AI generates the statement:

“I personally underwent laser treatment at this clinic and had incredible results.”

In reality, the influencer never received the treatment.

This is not merely inaccurate drafting.

It creates a false consumer testimonial.

If the advertisement also uses an AI-generated digital copy of a real individual, the specific 2026 prohibition concerning false AI-generated experiences or recommendations may additionally apply.

Businesses should therefore distinguish between:

a fictional advertising scenario

and

a fabricated personal experience presented as true.

The latter creates substantially greater regulatory risk.

12. What About the Social-Media Platform?

The fact that the advertisement appeared on Instagram, TikTok, YouTube or another platform does not automatically mean the platform becomes liable for every false statement uploaded by an advertiser.

Platform liability depends on the platform’s legal role, its involvement in the advertising process and the applicable regulatory framework.

However, Turkish consumer legislation expressly refers to media organisations (mecra kuruluşları) among the actors required to comply with advertising obligations under Article 61.

The legal analysis can therefore differ considerably between:

• a platform merely hosting user content;
• a platform selling and targeting advertising;
• an e-commerce marketplace creating its own campaign;
• and an intermediary itself acting as the advertiser.

For example, Ministry guidance concerning price advertisements recognises that an intermediary service provider may itself become an advertiser where it independently creates advertising concerning products sold through its platform.

The label “platform” therefore does not automatically answer the liability question.

The practical question is:

What role did the platform actually play in creating, controlling, targeting or publishing the commercial communication?

13. Is the AI Company Itself Responsible?

This is probably the most interesting question.

Suppose a brand uses a general-purpose generative AI service.

It enters the prompt:

“Write an advertisement for our vitamin product.”

The AI generates a false medical claim.

The company publishes it without reviewing it.

Would the consumer normally sue the AI provider instead of the advertiser?

Usually, the stronger consumer-law claim will be directed toward the commercial actors who used and published the advertisement.

A general-purpose AI provider is not automatically transformed into the advertiser merely because its software produced text used later by a company.

The situation could be different if the AI provider:

• directly operates an advertising service;
• designs and publishes the advertisement for the client;
• independently selects commercial claims;
• guarantees legal compliance;
• or otherwise assumes a direct role in the advertising transaction.

Contractual liability between the advertiser and AI provider may also arise depending on their agreement.

But from a compliance perspective, businesses should follow this principle:

Never assume that purchasing an AI service transfers the advertiser’s regulatory responsibility to the software supplier.

14. The Consumer May Have More Than a Regulatory Complaint

Misleading advertising may also affect the consumer’s contractual rights.

This is particularly important.

Law No. 6502 treats goods as defective where, among other things, they lack characteristics described in:

• advertisements;
• announcements;
• internet portals;
• labels;
• packaging;
• or promotional materials.

The Ministry of Trade confirms that products lacking one or more advertised characteristics may constitute defective goods under consumer law.

The same principle applies to services.

Article 13 of Law No. 6502 provides that a service may be defective where it does not possess characteristics communicated by the service provider or contained in advertisements or announcements.

This creates a direct connection between:

misleading AI advertising

and

consumer contractual remedies.

15. Example: AI Advertisement Promises a Feature the Product Does Not Have

Suppose a smartphone advertisement states:

“AI-powered instant Turkish translation available during every telephone call.”

The consumer purchases the phone specifically because of this feature.

The feature is unavailable in Turkish.

That scenario closely resembles a case already considered by the Turkish Advertising Board involving advertising of AI-related smartphone functionality. The Board found the advertisement misleading where a material limitation affecting consumer purchasing decisions was not disclosed.

From the consumer’s perspective, the problem may therefore operate on two levels:

Regulatory level

The advertisement may violate advertising law.

Contractual level

The product may fail to possess an advertised characteristic and therefore qualify as defective under Law No. 6502.

The consumer may consequently have access to the statutory remedies applicable to defective goods, depending on the circumstances.

16. The Same Rule Can Apply to AI-Generated Service Advertisements

Suppose a language school publishes an AI-generated advertisement stating:

“Every student reaches C1 level within three months.”

A consumer enrols because of the representation.

The school later says:

“That was just marketing language created by AI.”

That argument may not solve the problem.

Article 13 of Law No. 6502 provides that services lacking advertised characteristics can constitute defective services. The service provider is generally bound by advertised representations unless it can establish one of the statutory exceptions, such as lack of knowledge that could not reasonably have been expected, correction before the contract, or lack of causal connection between the advertisement and the purchasing decision.

The source of the marketing wording—human or machine—is not the decisive factor.

17. What Can the Consumer Claim?

The available remedies depend on the type of transaction and the loss suffered.

For defective goods, consumer legislation may provide remedies such as:

• rescission of the contract;
• replacement with a non-defective equivalent;
• price reduction;
• or free repair,

subject to the statutory conditions.

For defective services, Article 15 allows the consumer to choose, depending on the circumstances, among remedies including:

• re-performance of the service;
• free correction of the work resulting from the service;
• proportionate price reduction;
• or withdrawal from the contract.

Separate compensation may also become relevant where misleading conduct causes additional proven loss.

Accordingly, a successful Advertising Board complaint and an individual consumer claim are different remedies.

18. The Advertising Board Does Not Normally Calculate the Consumer’s Individual Damages

This distinction is practically important.

A consumer may complain to the Advertising Board about the misleading advertisement.

The Board examines regulatory compliance.

It may impose measures including:

• suspension of the advertisement;
• corrective measures;
• administrative fines;
• and, where statutory requirements are met, temporary suspension measures.

But the regulatory proceedings are not simply a substitute for a civil claim seeking repayment or damages.

A consumer who actually purchased the product because of the misleading advertisement may separately need to exercise consumer-law remedies before:

• a Consumer Arbitration Committee, where the monetary jurisdiction requirements are met; or
• the Consumer Court.

19. Administrative Fines Can Be Very Significant

Businesses should not view AI-advertising regulation as symbolic.

For 2026, the Ministry of Trade states that administrative fines for deceptive and misleading advertising and unfair commercial practices may fall between TRY 99,339 and TRY 39,916,524, depending on the applicable violation and circumstances. The statutory assessment may take into account matters including the nature of the violation, the benefit obtained or loss caused, fault and the economic position of the violating party.

Enforcement is also active.

The Ministry reported on 19 August 2026 that the Advertising Board had reviewed approximately 26,000 applications during the first seven months of 2026 and had imposed more than TRY 218 million in administrative penalties for misleading advertising and unfair commercial practices.

AI-generated marketing should therefore be treated as an actual regulatory-risk issue, not merely an ethical concern.

20. What If AI Creates a Fake Doctor?

This is an especially risky practice in health advertising.

Imagine a supplement advertisement showing an AI-generated person wearing a white coat.

The character says:

“As a cardiologist, I recommend this product to all of my patients.”

The character is entirely artificial.

Several legal questions arise:

• Does the advertisement clearly disclose that the person is AI-generated?
• Does the consumer reasonably believe the character is a real physician?
• Does the advertisement contain a health claim?
• Is that health claim independently permitted?
• Does it create false professional endorsement?
• Is a specialist title being used deceptively?

The 2026 amendments are particularly relevant because the new rules require disclosure for human-like AI characters and also prohibit misleading use of academic titles in advertising. The Ministry expressly announced these new protections when the amendments were adopted.

Simply inventing an artificial doctor therefore does not allow an advertiser to avoid the rules applicable to professional endorsements and health claims.

21. AI-Generated “Customer Reviews” Are Particularly Dangerous

Another emerging practice involves AI-generated reviews.

Examples include:

“★★★★★ Best product I have ever bought.”

“I used this for three weeks and my migraines disappeared.”

“Fastest delivery service in Türkiye.”

If these statements are generated by AI but presented as genuine consumers’ experiences, the advertisement may create a false impression about actual consumer experience.

Türkiye’s 2026 advertising reforms also strengthened the rules concerning consumer reviews. Reviews obtained from channels where the purchase process cannot be verified may not be published under the new rules.

Businesses should therefore not create fictional review profiles simply because generative AI makes them inexpensive and realistic.

22. The Brand Should Preserve the AI Prompt and Approval History

AI advertising also creates a new evidentiary problem.

Suppose the disputed advertisement states:

“Guaranteed 20% annual return.”

The company later argues that an external agency inserted that claim.

The agency says it came from the company’s prompt.

The AI system’s logs may show that the marketing manager specifically requested:

“Make the investment appear risk-free and guarantee returns.”

In future advertising disputes, evidence may therefore include:

• prompts;
• AI outputs;
• edit histories;
• campaign approval emails;
• agency instructions;
• internal compliance comments;
• publication records;
• model output versions;
• influencer scripts;
• and platform campaign settings.

Businesses should develop retention policies for significant AI-generated advertising campaigns.

23. Can Contract Clauses Transfer Responsibility to the Advertising Agency?

A contract may allocate financial risk between the parties.

For example:

“Agency shall indemnify the advertiser for penalties arising from claims created by the agency without client approval.”

Such a clause may matter in the internal relationship between the company and agency.

But it does not necessarily bind:

• the consumer;
• the Advertising Board;
• or other third parties.

A regulator may still identify the legally responsible actors under Law No. 6502.

The advertiser may subsequently pursue contractual recourse against the agency.

This distinction is critical:

External regulatory responsibility and internal contractual allocation of risk are not the same thing.

24. What If an Employee Publishes the AI Advertisement Without Approval?

Suppose an employee uses a generative AI system and posts a misleading advertisement directly through the company’s official Instagram account.

The board of directors had never approved the advertisement.

That may create internal employment or governance issues.

But the consumer sees an advertisement published under the company’s commercial identity.

The company should not assume that an internal authorisation failure automatically removes external legal responsibility.

Businesses using generative AI should therefore introduce controls governing:

• who may create advertising;
• who may approve it;
• who may publish it;
• what claims require legal review;
• and which sectors require specialist regulatory approval.

25. High-Risk Industries Need Stronger AI Advertising Controls

AI review should be particularly strict in regulated or high-risk sectors.

These include:

Healthcare

Claims relating to treatments, clinics, doctors, medical devices and health outcomes.

Food supplements

Statements concerning disease prevention, treatment, weight loss or health effects.

Finance and investment

Statements such as:

• guaranteed return;
• no risk;
• highest profit;
• government-backed;
• guaranteed investment.

Crypto assets

AI-generated statements concerning investment returns or regulatory status may create both advertising and capital-markets risk.

Cosmetics

Claims such as “clinically proven”, “permanent”, “medical result” or “doctor approved” require scrutiny.

Real estate

AI-generated photographs showing facilities, views or buildings that do not actually exist may mislead purchasers.

Automotive

Artificially created safety, performance or fuel-consumption claims may require objective proof.

26. A Practical Liability Map

When an AI-generated advertisement misleads a consumer, counsel should not ask only:

“Who used the AI?”

Instead, examine the complete advertising chain.

1. Advertiser / Brand

Usually the first and most important regulatory target because it commercially benefits from and publishes the claim.

2. Advertising Agency

May carry separate advertising-law responsibility where it creates or participates in the unlawful campaign.

3. Influencer

May be exposed where the influencer publishes misleading commercial representations, conceals the advertising relationship or fabricates experience.

4. Media Organisation / Relevant Platform

Its position depends on its actual legal and operational role in publishing or creating the advertisement.

5. Seller or Service Provider

May face direct consumer claims where the advertised product or service fails to possess promised characteristics.

6. AI Provider

Not automatically responsible merely because its general-purpose software generated the content. Liability requires a separate analysis of its actual role, contractual obligations and conduct.

7. Other Third Parties

Where deepfakes, expert endorsements, fabricated certifications or third-party identities are involved, additional civil, intellectual-property, personality-rights or unfair-competition claims may arise.

27. A Practical Example

Consider the following campaign.

A cosmetic brand asks an AI system:

“Create a persuasive advertisement for our hair serum.”

The system produces:

“Clinical studies show 93% of users regrow lost hair within two months.”

No clinical study exists.

The advertising agency creates an AI-generated female character indistinguishable from a real person.

She says:

“I personally used it and my hair completely grew back.”

The advertisement is sent to an influencer, who publishes it without an advertising disclosure.

A consumer purchases six bottles for TRY 18,000.

Who might face legal exposure?

Brand

For the unsupported product claim and the commercial advertisement.

Advertising agency

Because of its role in creating the misleading campaign.

Influencer

Because of the misleading promotional content and potentially inadequate commercial disclosure.

Seller/provider

Because the consumer may argue that the product lacks the characteristics represented in advertising.

Platform

Its responsibility would require analysis of its precise role.

AI provider

Not ordinarily the primary advertising-law defendant solely because the software generated the wording.

And there is an additional issue:

The photorealistic artificial spokesperson may itself trigger the 2026 AI disclosure requirement.

28. What Should a Consumer Preserve?

A consumer who believes they were misled by an AI-generated advertisement should preserve the evidence immediately.

The most useful evidence may include:

• screenshot of the advertisement;
• screen recording;
• URL or social-media account;
• publication date;
• influencer username;
• AI disclosure, if any;
• product page;
• invoice;
• payment receipt;
• product packaging;
• promotional email;
• WhatsApp messages;
• advertised characteristics;
• actual product characteristics;
• and evidence showing why the representation was false.

This is particularly important because digital advertisements can be changed within minutes.

The company may later replace:

“Guaranteed results”

with:

“Results may vary.”

The original advertisement must therefore be preserved.

29. What Should Companies Do Before Publishing AI Advertising?

A business using generative AI should establish an AI advertising review protocol.

Before publication, the following questions should be answered:

1. Did AI generate or materially alter the content?
2. Does the 2026 AI disclosure obligation apply?
3. Is any digital person likely to be mistaken for a real human?
4. Does the advertisement recreate a real person’s likeness or voice?
5. Does the advertisement imply real personal experience?
6. Can every objective product claim be substantiated?
7. Are health, financial or environmental claims involved?
8. Are comparison or superiority statements objectively supported?
9. Does an influencer need to disclose the commercial relationship?
10. Are testimonials genuine?
11. Are prices and discounts accurate?
12. Are material limitations clearly disclosed?
13. Does the final advertisement differ from the legally approved version?
14. Have prompts and approval records been preserved?

The rule should be simple:

No AI-generated factual claim should be published merely because it sounds convincing.

30. The Most Important Contractual Clause for Brands Using AI Agencies

Businesses should also update agency agreements.

An AI-advertising agreement should address at least:

• permitted AI systems;
• human review requirements;
• factual verification obligations;
• prohibited health and financial claims;
• intellectual-property risks;
• deepfake restrictions;
• use of real persons’ likenesses;
• influencer compliance;
• AI disclosure obligations;
• approval workflow;
• prompt and output retention;
• regulatory cooperation;
• notification of complaints;
• indemnification;
• and rights of recourse following regulatory penalties.

This does not eliminate statutory responsibility.

It helps determine who ultimately bears the financial consequences within the commercial relationship.

31. The Key Legal Principle: Responsibility Follows the Commercial Decision, Not the Algorithm

AI is changing how advertisements are created.

It has not eliminated the basic principles of advertising law.

If a company uses artificial intelligence to create a commercial message, decides to publish that message and economically benefits when consumers rely on it, it will be difficult for the company to argue that the algorithm alone should bear responsibility.

The Turkish framework reinforces this conclusion in three ways.

First, Law No. 6502 places the burden on advertisers to substantiate advertising claims and imposes compliance obligations on advertisers, agencies and media organisations.

Second, consumer-law provisions may make advertised characteristics legally relevant to whether the resulting product or service is defective.

Third, since 1 August 2026, Türkiye has had express rules requiring transparency for certain AI-generated advertisements and prohibiting false AI-generated digital copies of real people that appear to endorse or personally experience products.

Conclusion: “The AI Made the Mistake” Will Rarely End the Liability Analysis

When artificial intelligence generates a misleading advertisement, Turkish law does not simply search for the computer that generated the words.

It examines the commercial actors behind the advertisement.

The most important questions are:

Who commissioned the advertisement?

Who approved it?

Who published it?

Who benefited from it?

Who could substantiate the claim?

Who sold the resulting product or service?

Was AI use disclosed where required?

Was a synthetic person presented as real?

Was a real person’s digital likeness used to fabricate an endorsement?

In most ordinary cases, the brand cannot escape responsibility by pointing to the AI tool.

The agency cannot automatically escape by pointing to the brand.

The influencer cannot safely repeat claims merely because a brand supplied the script.

And a disclosure stating “AI-generated” cannot transform a false advertising claim into a true one.

For businesses, the safest principle is therefore:

Use AI to generate advertising ideas—not to replace legal and factual verification.

For consumers, the equally important principle is:

When an AI-generated advertisement causes a purchasing decision, preserve the advertisement itself.

Because once the misleading commercial representation can be proven, the dispute may no longer concern only advertising regulation. It may also become a consumer claim involving defective goods or services, refund rights, contractual remedies, compensation and—depending on the content—additional claims relating to unfair competition, personality rights or unauthorised digital impersonation.