The Ethical Case for Telling People When AI Is Involved

In late 2025, McDonald’s Netherlands released a holiday advertisement created entirely with generative AI without disclosing this fact. Social media identified it within hours, branded it “AI slop,” and the company withdrew it within three days. Simultaneously, Coca-Cola ran AI-generated holiday campaigns and openly communicated AI’s involvement through press materials and public channels. The reception was measurably more positive, despite both campaigns being equally controversial. The difference was transparency rather than creative quality.

This pattern occurs across industries and raises a fundamental question: when AI shapes information someone receives, influences decisions affecting them, or replaces expected human contact—do they deserve to know?

The answer is straightforward: yes.

The Gap Between Corporate Practice and Consumer Expectations

The disconnect between how organisations use AI and how they disclose it is striking. A 2025 survey by Forrester found that 71% of consumers want to know when AI is involved in customer-facing decisions. Yet only 28% of organisations actively disclose AI use beyond compliance requirements. When pressed, most companies cite fear of consumer backlash, concerns about appearing “cheap” or “impersonal,” or uncertainty about regulatory requirements.

This fear is understandable but misplaced. The McDonald’s example contradicts the assumption that disclosure automatically harms perception. What damages trust is discovery—when consumers realise they’ve been misled. The Coca-Cola example, by contrast, framed AI as a creative tool alongside human expertise. Transparency became part of the brand story.

In the DACH region specifically, this expectation runs deeper. Germany, Austria, and Switzerland have strong cultural emphases on disclosure, consent, and individual agency. The success of GDPR across the EU reflected not just legal compliance but genuine social expectation. Consumers in these markets expect to understand how their data is used, how decisions affecting them are made, and when automation is involved. An AI system making decisions about eligibility, pricing, or service quality without disclosure violates an implicit social contract in these markets.

The EU AI Act further codifies this. For high-risk systems—those used in employment, credit scoring, or public service delivery—documentation and explainability are mandatory. For systems generating synthetic content (text, images, video), disclosure is required “unless it is obvious in the circumstances.” That loophole creates ambiguity, but the intent is clear: citizens have a right to know when AI is shaping their information environment.

Understanding Disclosure: What It Means and Doesn’t

Disclosure is often misunderstood as a burden that requires lengthy explanations or technical detail. In practice, effective disclosure is simple, contextual, and proportionate.

What disclosure is:

A clear statement that AI was involved in creating or delivering the content
Explanation of why it was used (efficiency, scale, consistency, creativity)
Information about any limitations or uncertainties
Context on how the output was reviewed or assured

What disclosure is not:

A lengthy technical explanation of the model architecture
A disclaimer that absolves responsibility for output quality
An apology or justification for using technology
Acceptance of lower standards because “it’s AI”

In practice, disclosure can be as minimal as: “This email summary was generated using AI and reviewed by our team before sending.” Or: “These product recommendations are created by machine learning trained on customer behavior.” The statement acknowledges the technology without requiring the reader to become an expert in it.

The key principle is transparency of process, not transparency of mechanism. Consumers don’t need to understand how the language model works. They need to understand that a process different from human creation was involved.

The Front-Page Test

A useful heuristic for whether disclosure is needed: Would you be comfortable if this were on the front page of a major newspaper?

If a journalist discovered you were using AI in a customer-facing context and you’d scramble to explain or defend it, that’s a signal that disclosure should have happened upfront. The McDonald’s team presumably wouldn’t have been shocked by the discovery if they’d already told the market. Coca-Cola anticipated the conversation and shaped it.

The front-page test captures several underlying concerns:

Deception risk: Would consumers feel misled?
Trust impact: Would disclosure have built more trust than silence?
Reputational risk: Is there something about the use case that suggests embarrassment?

If the answer to any of these is “yes,” disclosure is necessary. And often, organisations discover that the actual concern isn’t about using AI—it’s about using AI poorly. A well-made AI-generated ad that people enjoy doesn’t need heavy disclosure. An obvious, low-effort output does, because it suggests corners were cut.

Disclosure as Starting Point, Not Endpoint

Disclosure alone doesn’t solve the ethical questions around AI use. It’s the foundation, not the solution.

Once a consumer knows AI is involved, several follow-up questions become relevant:

Who is accountable if the output is wrong or harmful?
How was the system trained and validated?
Are there alternatives if I don’t want AI involved?
How is my data being used?

For high-stakes decisions—credit approval, job screening, medical recommendations—disclosure must be paired with explainability. The consumer needs to understand not just that AI was involved, but how it reached that conclusion. The EU AI Act requires this for high-risk systems. But even for lower-stakes decisions, transparency of reasoning builds confidence.

A financial services company might disclose: “Your credit recommendation was generated by machine learning trained on 10 years of historical lending data. You can request an explanation of the specific factors that influenced your score.” That’s both transparent and empowering.

Practical Implementation Approaches

For organisations deciding how to disclose AI use, several approaches have proven effective:

Upfront disclosure. State AI involvement clearly in marketing materials, product descriptions, or service terms. Frame it positively (efficiency, consistency, scale) rather than defensively.

Contextual notation. Labeling in the moment: “Generated by AI” badges on content, “AI-assisted” in product descriptions, or “automated response” notifications in customer service.

Detailed documentation. For B2B and high-stakes contexts, provide full documentation: what data was used, how the model was trained, what testing was performed, what limitations exist.

Opt-in alternatives. Where feasible, offer consumers a choice. “Would you prefer AI-generated recommendations or suggestions from our team?” This respects agency and generates valuable feedback about preferences.

Regular auditing. Review AI outputs periodically for accuracy, bias, and appropriateness. Refresh disclosures if processes change. If audits reveal problems, take corrective action and communicate transparently.

The cost of these approaches is minimal compared to the reputational and legal risk of discovery. A brief disclosure statement costs almost nothing. Rebuilding trust after a discovered misuse costs significantly more.

Why DACH Markets Prioritize AI Transparency

The expectation for disclosure is particularly pronounced in German-speaking markets. This reflects several overlapping factors:

Legal tradition. Germany and Switzerland have strong legal frameworks around disclosure, consent, and individual rights. GDPR compliance in these markets reflects not just external regulation but internal cultural alignment. Consumers expect organisations to be transparent about processes that affect them.

Engineering culture. The DACH region has deep expertise in manufacturing, precision engineering, and technical excellence. There’s cultural expectation that complex systems are well-designed, well-documented, and rigorously tested. When AI is used poorly or without disclosure, it signals that corners were cut—a fundamental violation of quality expectations.

Social market orientation. Organisations in these markets are expected to balance profit with stakeholder responsibility. Transparency about AI use reflects concern for customer welfare and informed decision-making. It’s consistent with the region’s broader approach to corporate governance.

Regulatory environment. The EU AI Act, strengthened by German and Swiss input, sets disclosure requirements. Organisations in these markets are expected to lead, not merely comply. Early disclosure positions companies as responsible actors rather than reluctant followers of regulation.

In practice, this means DACH companies that adopt AI transparency early gain competitive advantage. They position themselves as trustworthy and responsible. Those that wait for regulation or grudgingly disclose after discovery damage their reputation among informed consumers and regulators.

Bottom Line

The ethical case for disclosing AI involvement is straightforward:

Consumers expect transparency when AI shapes their information or decisions
Disclosure, when done well, builds trust rather than damaging it
Discovery without prior disclosure damages reputation significantly
The cost of transparent disclosure is minimal
The legal environment (particularly EU AI Act) increasingly requires it

For organisations operating in DACH markets, AI transparency should be standard practice, not a compliance checkbox. It reflects the region’s cultural values around disclosure, consent, and individual agency. It aligns with emerging legal requirements. And it makes business sense: transparent organisations build trust more effectively than those caught using AI without disclosure.

The choice between the McDonald’s and Coca-Cola approaches isn’t really about how good the AI output is. It’s about respect for the consumer’s right to understand how they’re being influenced. That respect, clearly communicated, is where trust begins.

This article is based on extensive research. The full evidence base with sources is available upon request.