For years, businesses have been told that AI regulation is coming.

In Europe, part of it has now arrived.

On August 2, 2026, major parts of the European Union’s AI Act became applicable. The European Commission and national authorities also began enforcing new transparency requirements covering certain AI interactions and AI-generated or altered content. 

This is not only an issue for technology companies. Marketing teams use AI to operate chatbots, generate images, edit videos, draft articles, personalize campaigns and automate customer conversations. Some of those activities may now carry specific transparency responsibilities.

The message for business leaders is clear: AI governance can no longer sit entirely with IT. Marketing needs a seat at the table because it controls many of the places where customers encounter AI.

The Rules Can Reach Beyond Europe

A business does not necessarily need a European headquarters to fall within the AI Act’s scope.

According to the European Commission, the framework applies to public and private organizations inside and outside the EU when they place an AI system or general-purpose AI model on the EU market, put one into service or use it in the EU. 

That matters for companies running international websites, selling to European customers or using AI-powered experiences in multiple markets.

It does not mean every organization using an AI writing assistant automatically faces the same obligations. The Act follows a risk-based approach, and the requirements depend on the system, its purpose and how it is used.

Businesses should not guess which category applies. The first practical step is understanding where AI appears in customer-facing and employee-facing processes.

Transparency Is Now an Operational Requirement

The new rules focus partly on situations in which people need to know that AI is involved.

The European Commission says chatbots and other interactive AI systems must inform users when they are dealing with AI rather than a person. Deepfake images, audio and video must be labeled, while certain AI-generated or altered content must include machine-readable marks that make its artificial origin easier to detect. 

The detailed requirements contain exceptions and technical distinctions. Not every piece of AI-assisted copy needs a large warning attached to it.

However, a marketing team should be able to answer basic questions about every AI-supported experience:

• Is the customer interacting with AI or a person?
• Could an image, voice or video be mistaken for a real event?
• Does the content make a claim about a real customer, employee or public figure?
• Is the material intended to inform the public about a matter of public interest?
• Does the platform or tool preserve machine-readable information about the content’s origin?

If nobody owns those questions, compliance can easily become an afterthought added just before publication.

Start With an AI Marketing Inventory

Many businesses cannot govern their AI use because they do not have a complete picture of it.

The website team may have installed a chatbot. The social team may be generating campaign images. An outside agency may use AI to create video variations. Individual employees may be using their own tools to draft copy or analyze customer information.

Create a simple inventory that records:

1. The AI tool or feature being used
2. The marketing activity it supports
3. The data entered into it
4. The audience that sees the result
5. Whether the output is reviewed by a person
6. Whether an AI disclosure or label is present
7. The employee or partner responsible for the activity

Do not limit this review to tools labeled as “AI platforms.” AI features are increasingly built into advertising platforms, design software, customer service systems, analytics products and CRM tools.

The objective is not to stop useful experimentation. It is to make sure the business knows where AI is operating under its name.

Treat Your Chatbot Like a Customer Touchpoint

A chatbot should not be treated as a technical widget hidden in the corner of a website. It is part of the customer experience.

If a visitor believes they are speaking with an employee, the interaction begins with a false assumption. A clear introduction such as “You’re chatting with our AI assistant” can set the right expectation before the conversation continues.

Transparency is only the beginning. The team should also decide what the chatbot may answer, when it must transfer someone to a person and what information it is allowed to collect.

Review its most common conversations. Look for inaccurate answers, unsupported promises and situations where customers repeatedly ask for clarification.

The best chatbot is not necessarily the one that handles the most conversations without human help. It is the one that helps customers reach the right next step without creating confusion.

Build Disclosure Into the Creative Workflow

AI-generated media can move quickly from experiment to campaign asset. That speed makes it easy to overlook how the finished work will be interpreted.

Add a disclosure review to the same process used for brand, legal and factual approval. The person approving the asset should know whether AI was used to generate or materially alter it.

The review should consider the overall impression, not only the caption.

A generated lifestyle image may be understood as a creative illustration. A realistic video of a person saying words they never spoke presents a different level of risk. An altered product demonstration may also mislead customers if it shows a feature or result they will not receive.

The European Commission’s Article 50 guidance, published on July 20, 2026, was designed to help providers and deployers apply the transparency obligations consistently and proportionately. 

Marketing teams should use the official guidance alongside advice from qualified legal or compliance professionals when determining what a particular campaign requires.

Ask Vendors Better Questions

Using a third-party platform does not remove the need to understand what it does.

Before adopting an AI marketing tool, ask the vendor:

• Does the tool add machine-readable markers to generated content?
• Are those markers preserved when files are downloaded, edited or resized?
• How does the product notify users that an interaction is AI-powered?
• What customer data is stored, and for how long?
• Is submitted data used to train or improve the vendor’s models?
• What documentation is available to support regulatory review?

Keep the answers with the contract and implementation records. If the tool or its policies change, review the use case again.

A feature that was appropriate when it only suggested copy may require a different approval process once it begins publishing content or interacting directly with customers.

Governance Should Help Marketing Move Faster

Good governance is often mistaken for a long policy document that slows everyone down. In practice, uncertainty creates more friction.

When marketers know which tools are approved, what data they may use, when disclosure is needed and who can approve an unusual campaign, they can work with greater confidence.

Create a short operating guide with examples drawn from your actual marketing. Train employees and agency partners on it. Review it as tools, campaigns and regulations change.

The AI Act is a European law, but the operational lesson travels well: customers, regulators and business partners increasingly expect companies to know when AI is being used and who is accountable for the result.

AI adoption has moved faster than many internal processes. Now is the time to close that gap.

Need a clearer marketing strategy for a changing AI landscape? Talk to The Go Agency about building campaigns that use new technology without losing sight of your audience, brand or business goals.