AI for business · Industry guide

📣 AI rules: Marketing, media and design

Content is where transparency rules apply: deepfakes must be labelled, Slovenian media must label all AI content, and copyright questions arise with every campaign.

Typical AI uses and their risk level

Minimal
Drafting texts, slogans and social posts

Minimal risk; a person reviews before publishing.

Transparency
Realistic AI images or video of people or places

Deepfake label required since 2 Aug 2026.

Transparency
AI voice clone in an ad

Label it; you need consent of the person whose voice is cloned.

Transparency
AI-written news on a portal

ZMed-1: label at beginning and end; AI Act for public-interest texts.

Minimal
Personalised product recommendations

Minimal under the AI Act; GDPR consent/legitimate interest and DSA transparency for platforms.

What to do

  • Create a simple labelling rule: which content gets which label.
  • Keep records of prompts and human edits for key creative assets.
  • Check licences of AI tools for commercial use and indemnities.
  • Never use a real person’s face or voice without consent.

Common pitfalls

  • AI-generated fake reviews or testimonials are banned consumer practices.
  • Political campaigns: extra transparency and targeting rules since Oct 2025.

Examples

A tourism board uses an AI photo of a sunset over Bled

Realistic image of a real place → label “AI-generated image”. A clearly stylised illustration needs only a light mention.

An agency designs a logo with Midjourney

Copyright may not arise without human creative input — rework it and register a trademark.

General information, not legal advice — consult a lawyer for borderline cases. Regulations reviewed on 29 Sep 2026.