martin@tschapke.com LinkedIn Legal notice Privacy EN / DE
MRT Martin Reza Tschapke

Article

AI is booming in mid-sized firms, but 84 percent of trade businesses aren’t considering it

One in four companies in Germany now uses AI, while for most trade businesses it’s still far away. In five charts, I show where AI stands today, what holds companies back and where small businesses can sensibly start.

Oct 6, 2026 · 3 min read
by Martin Reza Tschapke
Two people in a workshop pointing at a laptop showing a process diagram

There’s a lot of talk about artificial intelligence. How many businesses actually use it, and for what, is shown by official statistics and several large surveys. The result: use is growing fast, but very unevenly.

One in four companies uses AI

According to the Federal Statistical Office, 26 % of companies with 10 or more employees used AI in 2025, up from 12 % in 2023. That’s more than double, with growth in every size class. At the same time, the gap between small and large companies has widened: 23 % of companies with 10 to 49 employees use AI, compared with 57 % of those with 250 or more.

Three small column charts: share of companies using AI in 2023, 2024 and 2025 by size. 10 to 49 employees 10, 17 and 23 percent. 50 to 249 employees 16, 28 and 36 percent. 250 or more employees 35, 48 and 57 percent.
Companies with 10 or more employees. Since 2025, the Federal Statistical Office counts companies according to the EU definition and asks about one more technology, so the jump isn’t only growth.

Among companies using AI, 52 % use it to analyze text, and just as many to generate images, video or audio. Speech recognition (42 %) and generating text or code (35 %) follow. Generative AI has arrived in everyday office work.

In a European comparison, Germany ranks 8th out of 27, well above the EU average of 20 %. Denmark leads (42 %), followed by Finland and Sweden.

Bar chart: share of companies using AI in the 27 EU countries in 2025. Denmark 42 percent, Finland 37.8, Sweden 35, Germany in 8th place with 26 percent, EU average 20 percent, Romania 5.2 percent.
Companies with 10 or more employees, 2025.

In the skilled trades, AI is still far away for most

The skilled trades look very different. In a Bitkom survey of 504 trade businesses, only 4 % used AI in 2025 and 9 % were planning to. For 84 %, AI wasn’t on the agenda. That fits another finding of the same study: 72 % of businesses say they’re too busy to deal with digitalization. And only 29 % have staff who know how to work with AI.

That it doesn’t have to be about size is shown by KfW, the German development bank. Among small and mid-sized businesses, 19 % of those with fewer than 5 employees used AI between 2022 and 2024. According to KfW, digitalization and a culture of innovation matter more than headcount. In construction, however, the share was only 8 %.

What AI is used for

Companies that use AI mainly apply it where a lot of writing and communication happens. In the Bitkom study 2026, 72 % of companies with AI name customer contact, 54 % marketing and communication. Controlling, production and knowledge management follow far behind.

Bubble chart: areas where companies use AI in 2026. Customer contact 72 percent, marketing and communication 54, controlling 25, production 22, knowledge management 22, own products 20, research 16, sales 14 percent.
Companies with 20 or more employees that use AI (343 answers), multiple answers.

What holds companies back: first knowledge, then data protection

What’s interesting is how the hurdles shift. Companies that don’t use AI yet mainly name a lack of know-how (85 %). Among those already using AI, only 25 % see this hurdle, and data protection becomes the biggest one (66 %).

Gap chart: hurdles to AI for companies that use AI and companies that don’t. Lack of know-how 25 vs. 85 percent, legal uncertainty 56 vs. 66, costs hard to calculate 54 vs. 63, staff 43 vs. 51, data protection 66 vs. 49, missing data 19 vs. 40, quality 26 vs. 40, no use cases 11 vs. 30 percent.
Companies with 20 or more employees, 2026.

On the legal side, there’s some relief. From February 2025, the EU AI Act required companies to do their best to ensure sufficient AI literacy among their staff. Since 27 July 2026, a weaker version applies: businesses take measures to support this literacy, but don’t have to guarantee a particular level. No fine is provided for it. Anyone using chatbots, however, has had to make it recognizable since August 2026 that people are talking to an AI.

Where small businesses can start

For small businesses, AI doesn’t have to be a major project. Three entry points save time quickly, and there are already figures for all three:

Three cards with entry points into AI for small businesses: texts and quotes, customer communication, and bookkeeping and administration, each with an example and a key figure.
Key figures from KfW, Bitkom and Eurostat with different populations.

Where to start

  1. Pick one task. Something that comes up every week and involves a lot of writing, such as quotes or replies to inquiries.
  2. Set rules. No customer data in free tools, always check the results, and everyone knows which tool is allowed. That’s also the simplest contribution to AI literacy in your team.
  3. Build it into existing processes. AI helps most where it’s integrated into your CRM, email and website, rather than running in an extra window.

I use AI every day for texts, analyses and automations, for example in the CRM systems I build for businesses. If you want to know where AI could save time in your business, get in touch.

Sources

  1. Federal Statistical Office (2023, 2024): press releases no. 453 of 27 November 2023 and no. 444 of 25 November 2024 on AI use in companies (in German). destatis.de
  2. Federal Statistical Office (2025): companies using artificial intelligence, table 2025 (in German). destatis.de
  3. Eurostat (2025): Use of artificial intelligence in enterprises (isoc_eb_ai). ec.europa.eu/eurostat
  4. Bitkom (2026): AI in the economy 2026, press release of 14 September 2026, 603 companies with 20 or more employees (in German). bitkom.org
  5. Bitkom (2026): Digitalisierung des Handwerks (digitalization of the skilled trades), study report, 504 businesses (in German). bitkom.org
  6. KfW Research (2026): KfW Fokus Volkswirtschaft no. 533 and no. 554 on AI in small and mid-sized businesses (in German). kfw.de
  7. Regulation (EU) 2024/1689 (AI Act). eur-lex.europa.eu
  8. Regulation (EU) 2026/1744 (Digital Omnibus on AI), Official Journal of 24 July 2026. eur-lex.europa.eu

More articles