Use AI for · Marketing · 26 Sep 2026
Use AI for
AI in marketing

AI in marketing:more ideas and variants,your brand stays yours

AI drafts copy, images, videos and ad variants in a fraction of the usual time. This guide covers which marketing tasks it suits, how to go about it, where it tends to go wrong and what copyright, data protection and the EU AI Act expect from you.

Get in touch
On this page
Enquiry
Nikolai Schöbel und Jeremias Burger, Co-Founder Scalableloops

Let's talk about your project.

First we check whether the project fits your business model. Then you get a proposal with phases and effort.

Discuss AI in marketing or call: +49 151 1576 5566
Use AI for · Marketing

AI in marketing means using language and image models for copy, images, videos, ads and analysis so your team gets to drafts and variants faster. The AI supplies raw material and ideas, while decisions about message, brand and facts stay with you. To use it safely, settle three things first: what data may go in, who checks results before publication and when AI content has to be disclosed.

In brief
  • AI works well for drafting copy, images and videos, for ad variants and for analysing customer feedback.
  • Good results need a prompt that names audience, tone, channel and verified facts about your offer.
  • According to Germany’s Federal Ministry of Justice, purely AI-generated images are not protected by copyright.
  • Since 2 August 2026, Article 50 of the EU AI Act requires deepfakes to be disclosed as artificially generated.
Published 26 Sep 2026Nikolai Schöbel and Jeremias Burger10 min read
Nikolai SchöbelJeremias Burger

Nikolai Schöbel and Jeremias Burger

Co-founders of Scalableloops GmbH. Nikolai Schöbel leads online marketing and AI strategy, Jeremias Burger the AI architecture. Both build AI systems and train teams on them in their own agency work.

On this page
  1. What does AI bring to marketing?
  2. Which marketing tasks is AI suited to?
  3. How do you use AI in marketing step by step?
  4. What does a good prompt for marketing copy look like?
  5. How can AI help with content that people find in search?
  6. What should you watch for when AI writes your ads?
  7. Who owns AI images, and when do you have to disclose them?
  8. How do you protect customer data and prepare your team?
  9. Frequently asked questions
  10. How to get started with AI in marketing
  11. Where the information on this page comes from
Overview

What does AI bring to marketing?

Above all, AI in marketing gets your team past the blank page. Instead of starting from nothing, you get several drafts for a post, a subject line or an ad visual within minutes, then choose, refine and test. Speed goes up, and so does the need for careful review.

AI has reached the majority of companies. According to a Bitkom survey of 603 German companies with 20 or more employees, published in September 2026, 57 percent use AI, up from 36 percent a year earlier. Marketing is a natural place to start because it produces a lot of text, a lot of images and a lot of variants.

What AI does not bring is knowledge of your offer, your customers and your brand. Without that input it writes in generalities, tends to exaggerate and, when in doubt, invents features your product does not have. How useful it is depends on how well you brief it and how carefully you check the result.

Tasks

Which marketing tasks is AI suited to?

AI is strongest where many drafts or variants are needed and a person makes the final choice. The overview shows typical tasks, what AI delivers and what stays with you.

For detailed guidance on each task, see our guides to using AI for writing, using AI for image creation and using AI for video creation.

TaskWhat AI deliversWhat stays with youType of tool
Social media postsDrafts, variants for different channels, topic ideasSelection, brand voice, facts, sign-offAI chats such as ChatGPT, Claude, Gemini
Newsletters and emailsSubject lines, structure, first draftOffer, audience, legal points, sendingAI chats, AI features in your email tool
Images and visualsDrafts, backgrounds, format variantsBrand, rights, disclosure, final approvalImage generators such as Adobe Firefly or Midjourney
VideosScripts, editing ideas, short clipsMessage, real footage, sign-offAI chats for scripts, video generators
AdsHeadlines and descriptions in variantsKeywords, budget, checking every variantAI chats, AI features of the ad platform
AnalysisSummaries of reviews and surveys, patterns in feedbackInterpretation, decision, actionAI chats on a business account
Process

How do you use AI in marketing step by step?

The most common mistake is sending AI off without any groundwork. This process makes sure results sound like your brand and only contain claims you can back up.

  1. 01

    01

    Gather the basics

    Write down who your audience is, how your brand speaks, which words you avoid and which facts about your offer are verified. Give this brief to the AI with every task.

  2. 02

    02

    Keep the task narrow

    One post for one channel with one goal gets better results than “do our marketing”. State the channel, length, goal and the action you want from the reader.

  3. 03

    03

    Generate variants

    Ask for several drafts or designs and choose between them. Request genuinely different approaches rather than small tweaks of the same sentence.

  4. 04

    04

    Check the facts

    Compare every figure, product feature and promise with your own documents. Anything you cannot back up comes out.

  5. 05

    05

    Approve and disclose

    A person approves what gets published and checks whether images, videos or text need to be disclosed as AI-generated.

  6. 06

    06

    Measure and refine

    Review which variants work and feed what you learn back into your brief for the next round.

Prompts

What does a good prompt for marketing copy look like?

A good prompt names audience, channel, tone, goal and the facts the copy may rely on. For a LinkedIn post it might read: “Write a LinkedIn post about our new maintenance service for commercial refrigeration. Audience: operations managers in food production. Tone: factual, no superlatives. Use only these facts: [list]. Do not invent any figures. End with an invitation to get in touch. Give me three different openings.”

For ads, a clear brief on length and content helps: “Write ten headlines of no more than 30 characters for a search ad on the keyword ‘solar panel maintenance’. Each headline should name a different benefit from this list: [list]. No exclamation marks and no promises that are not on the list.”

To analyse customer feedback, try: “Here are anonymised reviews from our customers. Summarise what is often praised and what is often criticised, and back up each point with a quote from the reviews.” Remove names and anything that could identify individuals first.

It is worth ending every prompt with a request for self-checking: “Highlight every statement in your text that did not come from the information I gave you.” That quickly shows you where the AI has added something you need to check or delete.

Search

How can AI help with content that people find in search?

Google judges content by its quality, not by how it was produced. That is how Google put it on its Search Central blog in February 2023. At the same time, its spam policies prohibit producing many pages mainly to manipulate rankings without helping users, and they say explicitly that this applies whether the pages come from people, automation or both.

For your marketing, this means AI can help with writing as long as the content contains first-hand experience, real examples and verifiable statements. Mass-produced pages with interchangeable content are a risk. Our guide to AI SEO explains what matters when you use AI for search engine optimisation.

More and more people put their questions straight to AI assistants. For your brand to be mentioned there, you need clear answers, facts with sources and a clean structure. We explain the terms behind this in our article on GEO, AEO and LLMO.

Ads

What should you watch for when AI writes your ads?

Ad platforms now build AI in themselves. In Google Ads, text customisation for search ads can generate additional headlines and descriptions. According to Google, it combines snippets from your landing page titles and descriptions with generative AI. You can switch the feature on or off for each campaign, and the generated text can be reviewed in the asset details report.

Features like this save work, but they put copy into circulation that nobody on your team has read beforehand. Decide deliberately whether to use them and check regularly what is being served. Google itself recommends keeping the titles, descriptions and meta tags of your landing pages up to date, because the ads draw on them.

You can now also advertise inside AI assistants themselves. Our ChatGPT ads guide covers what to bear in mind.

Images and law

Who owns AI images, and when do you have to disclose them?

Germany’s Federal Ministry of Justice takes the view that copyright only protects personal intellectual creations by a human being, so purely AI-generated content has no copyright protection. For your marketing, this means competitors can reuse a purely generated visual without you being able to rely on copyright. Protection only becomes possible where your own creative contribution shapes the result.

Conversely, an AI image can infringe other people’s rights, for example if it resembles a well-known character, a trade mark or a real person. Check visuals for this before publishing. Some providers highlight their training data: Adobe states that its Firefly models were trained on licensed content such as Adobe Stock and on public domain content, and not on its users’ content.

Article 50 of the EU AI Act has applied since 2 August 2026. Under paragraph 4, deployers must disclose that image, audio or video content has been artificially generated or manipulated if it is a deepfake, meaning content that resembles real people, places, objects or events and could falsely appear authentic. For evidently artistic, creative or satirical works, an appropriate notice is enough. A photorealistic AI image of a person in an ad should therefore be checked against the disclosure obligation before it goes live. The transition period until 2 December 2026 applies only to machine-readable marking by providers of generative systems, not to disclosing deepfakes.

Data and team

How do you protect customer data and prepare your team?

Customer lists, orders and reviews with names are personal data. They only belong in AI tools covered by a data processing agreement under Article 28 GDPR and where training on your inputs is excluded. The German data protection authorities (Datenschutzkonferenz) recommend that employers provide work accounts rather than letting staff use personal ones. For many marketing tasks, anonymised or aggregated data is enough anyway.

Under Article 4 of the EU AI Act, organisations that use AI have had to take measures to support AI literacy since 2 February 2025. Since 27 July 2026, Article 4 has applied as amended by Regulation (EU) 2026/1744: you do not have to guarantee a specific level of competence for any individual, and no certificate is required. For a marketing team, this means anyone working with AI knows the limits of the tools, what data may go in and the rules on disclosure. More in our article on the AI literacy obligation under Article 4.

Put your rules in writing: which tools are allowed, who signs off, when content is disclosed and where the brand and fact documents are kept. That saves questions and means new colleagues work to the same rules from day one.

Frequently asked questions

Frequently asked questions

Which AI tools are useful for marketing?

For copy and ideas, AI chats such as ChatGPT, Claude or Gemini; for images, generators such as Adobe Firefly or Midjourney. On top of that come the AI features built into ad platforms and email tools. What matters more than the individual tool is a business account with a contract and a clear sign-off rule.

Does Google penalise AI-written content?

Not because of how it was made. Google has said it rewards content based on quality, however it is produced. Its spam policies do prohibit producing many pages mainly to manipulate rankings without helping users, regardless of whether a person or a machine wrote them.

Can I use AI images in advertising?

Yes, provided you respect third-party rights and check the disclosure rules. According to Germany’s Federal Ministry of Justice, purely AI-generated images are not protected by copyright. If an image looks like a real photo of an actual person, place or event, Article 50 of the EU AI Act may require disclosure.

Do I have to label AI-written marketing copy?

There is no general labelling requirement for advertising copy. Article 50(4) of the EU AI Act requires disclosure for AI-generated text published to inform the public on matters of public interest. That does not apply if a person has reviewed the text and someone holds editorial responsibility.

Can I enter customer data into ChatGPT?

Not on a personal account. For personal data you need a business account with a data processing agreement and training switched off. Often it is enough to anonymise or aggregate the data first. Clarify the use case with your data protection officer.

Will AI replace a marketing agency or an in-house team?

AI speeds up drafts and variants. Strategy, knowing your customers, brand management and checking facts and rights remain human tasks. Teams gain most when they write down what they know about their offer and audience and give it to the AI.

What now

How to get started with AI in marketing

  1. 01

    Pick your tasks

    Choose two or three recurring marketing tasks that involve a lot of text or many variants, such as posts, newsletters or ad copy.

  2. 02

    Write the brief

    Record your audience, brand voice, words to avoid and verified facts about your offer in one document you give to the AI.

  3. 03

    Set the rules

    Settle tools, business accounts, data processing, sign-off and disclosure before the first AI result is published.

  4. 04

    Build routines

    Practise with your team on real tasks and turn recurring steps into routines. We can help with an AI agent system set up for your business.

AI in marketing brings speed and variety to your drafts. What makes the difference is the brief you give it and the care with which a person checks facts, rights and brand before anything is published.

or call: +49 151 1576 5566

Further reading
Sources

Where the information on this page comes from

Projekt-Detail

Problem

Our solution

Our role

Highlights

    Got a project in mind?

    We reply personally. First a use-case check, then an architecture proposal.

    Start your inquiry