Where I use AI in marketing, where I don’t, and why

After publishing my blog about “buying a house using ChatGPT” last year, a lot has changed. AI has improved significantly, people are more familiar with the dos and don’ts, and agents are now all the rage. While finalizing the renovation of my house, I decided this time to write about how I personally use AI today in my job and for which parts of marketing I don’t like to use it.

The gap in marketers’ AI skills is widening fast, and it will definitely affect the already fierce marketing job market. While some still mainly use AI for translation or copywriting, others are already running fully AI-driven systems that automate the entire marketing process from data and creative to execution and optimization, and use AI agents on top of CDPs (customer data platforms) to turn natural language into insights, segments, and continuous, closed-loop growth. The development of marketing practices is exciting, but it’s also moving very fast. It’s always scary to think about your own field going through a massive transformation. However, staying on top of developments, being aware of new possibilities, and testing new methods will help a lot.

Due to the mainstream adoption of AI, the content creation process also looks a lot different now for many marketers. Have you heard of the dead internet conspiracy? It suggests that a large part of the internet is nowadays not even made by real people, but content, comments, followers, customer reviews, and other engagement that seem human are actually AI-generated or bots. A rather innocent example: a “co-player” of mine in a tower defense mobile game actually turned out to be a bot, as I was told by the developer who couldn't stay silent anymore after listening to me hyping about the player being so friendly. Bots have their place in games, but wouldn’t it be a bit spooky if it turned out none of your favorite blogs, product pages, news stories, guides, or manuscripts for YouTube videos were written by humans? That said, it’s worth taking all of this with a pinch of salt. Cyber security expert and ethical hacker Laura Kankaala speaks about an ever-changing internet rather than a dead one in this podcast episode on YLE Areena (in Finnish).

In my own work, there are areas where I wouldn’t go back to the “old ways”, and others where I still prefer them. Let’s take a look at both.

Where I like to use AI in marketing

1. Improve my craft

I use AI to clarify my message, correct language or grammar, and refine text. Sometimes we overcomplicate messaging that’s supposed to hook potential customers. Your brand message should clearly and briefly explain what you offer and why you’re the right choice. This can be surprisingly hard, but AI is very good at shortening and simplifying messaging so customers get it instantly. It’s never the customer’s job to figure out on their own what you sell or why you’re the best.

2. Summarize meeting recordings

This assists me with internal communications or turning a webinar recording into a short written summary that I can use to promote the webinar as on-demand. Another example is identifying which topics resonate most with your audience based on discussions and interactions. It goes without saying, but ask permission before recording anyone.

3. Use transcripts for content creation

I can also use transcripts of some of our internal meetings to create thought leadership content. A common and fully understandable challenge for marketing is the availability of your subject matter experts for new content creation. In consulting, experts are sometimes fully booked with client work and simply don’t have time for marketing activities. One solution to that is to record your internal meetings (with permission from the speakers, of course). Identify topics that would interest people outside your organization, record, transcribe using AI, and use the transcription to create a blog or a white paper.

4. Analyze large amounts of information

This could be your internal materials like notes, sales decks, and proposals. Pattern recognition is one of AI’s biggest strengths, and it can do this very quickly. I use this to quickly gather all material we have related to a specific use case or industry, and structure it into a presentation. Pro tip: check with your administrator before connecting AI (agents) to internal systems.

Photo from Snowflake's Data for Breakfast in Helsinki, March 2026.

Where I prefer not to use AI in marketing

1. Short-form content / messages

Why not spend 5–15 minutes writing a message that reflects what you know about the topic, and what you know about your audience? Perfectly curated, grammatically impeccable, and in some cases soulless copywriting is a bit boring. I think people are increasingly suffering from “perfection fatigue” on social media and that explains why the so-called “lo-fi” content is winning over viewers right now across platforms. 

Sometimes, it seems that as we get more efficient, we don’t actually gain time, we just get even more stressed as we understand how much more we could do if we were just a little bit more efficient. That’s when AI slop starts to feel tempting, because “you can generate so much content in so little time, which is perfect for your SEO or social media strategy”. I have tested a lot of different methods, from generating all my social media posts with AI, to doing a mix, and came to the conclusion I like my own writing the most, even though it's far from perfect. 

There’s a saying “the less you do, the more you get done", and I think it’s all about doing stuff thoroughly instead of doing a million things all at once and none of them properly.

2. Create images

There are very talented photographers, graphic designers and artists out there that deserve to have their work seen. And your marketing deserves high-quality material with some thought behind it. I think AI is useful for ideation, but final visuals still benefit from human intent and craft. Cost is one argument, but there are a lot of human-made photography and visual design elements available for free on e.g. Unsplash that can be used for most commercial, personal projects, and for editorial use. It’d be a bit unsettling to see all humans in ads being replaced by AI-generated characters.

3. Outsource your thinking or imagination

Which one is more fun and actually develops your work relationships and social skills: brainstorming ideas (for events, content, etc.) with your colleagues, or asking AI to shoot ideas your way? It’s tempting to outsource every small task or idea to AI, but in doing so, we stop exercising our own critical thinking. I use my phone as a "dictaphone" to record all my ideas when I'm not in position to write them down, e.g. during dog walks, and have a spreadsheet with all my ideas (some good, some shitty) and then pitch them to my colleagues. 

To be transparent, I used AI to correct the language of this blog, as well as to refine/clarify some sentences. All opinions are my own, and I also do not intend to AI-shame anyone; you should go with what works best for you and I’ll do the same. 

Are you interested in advancing your own AI usage?

We’re soon launching a new guide on how to build your own agentic workflows for roles such as marketing, sales, strategy, HR, etc., to show how you can go beyond simple chat and copy-pasting when using AI. Keep an eye out for it!