AI bootcamp for EU affairs, day 7: create better images with AI
Image generation is easy now. The brief is the skill. Here is the 2026 landscape, five ingredients that work in almost any tool, and the checks a campaign owes its audience before publishing.
Sebastián Rodríguez Pérez is a European campaign strategist with direct experience running and advising campaigns for some of Europe's largest pro-European organisations. He founded the European Campaign Playbook to codify practitioner knowledge across the EU's 27 political environments and make it accessible to campaign professionals, advocacy teams, and civic organisations working at the European level. His work spans electoral strategy, EU public affairs, digital campaigning, and the integration of AI tools into political communication workflows.
Disclosure
Sebastián Rodríguez Pérez is the founder of the European Campaign Playbook and Campaign Intelligence Library. He has worked with pro-European organisations across multiple EU member states. He maintains editorial non-partisanship in all published content; where articles relate to campaigns or organisations he has directly worked with, this is disclosed within the article. He holds no current positions in active electoral campaigns.
What it covers
What the AI image landscape looks like in 2026, how to pick a tool, how to write an image prompt as a visual brief, how to improve a picture without starting again, how to use reference images, and the checks EU public affairs teams should make before publishing synthetic visuals.
Who it is for
Public affairs, advocacy, policy and communications teams producing campaign visuals without an in-house designer, and anyone introducing AI into an existing creative workflow.
When to use it
After you know what the campaign is trying to say. If the message is still unclear, go back to the strategy: faster images will not fix it.
Key takeaway
The model matters and the brief matters more. Describe what should be visible, how it should look, how it should be composed, what atmosphere it should have and what must stay out. Then change one variable at a time, finish the asset in your own design tools, and review it before it goes anywhere.
In brief
- AI image tools now split into three groups: conversational assistants, design-led products, and specialist or open models for production.
- Start with the image capability inside the AI platform your organisation already approves, and move on when you have a specific reason.
- Five ingredients cover most briefs: subject, style, framing, light and colour, mood.
- Constraints are the sixth: say what must stay out, and what must stay unchanged.
- Generate three genuinely different directions before polishing one.
- Refine one variable at a time, so you learn which instruction did the work.
- Use AI to prototype and produce components, then apply real branding, typography, accessibility and editorial review.
- Photorealistic synthetic content deserves extra scrutiny, because viewers read photographs as evidence.
- EU AI transparency rules and the political-advertising regulation can add obligations, and they are two separate reviews.
- The Commission publishes free EU icons for labelling AI content. Using them is optional; the Article 50 duty behind them is not.
Framework
The five ingredients of an image brief
Answer these five questions and you have a brief. Add the constraints and you have a usable one.
| Layer | Label | Description |
|---|---|---|
| What should we see? | Subject | Volunteers planting young trees in a neighbourhood park, with spades, gloves and watering cans. |
| What visual language? | Style | Realistic editorial photography, closer to community-event photography than polished stock. |
| How is it composed? | Framing | Medium-wide, planting in the centre-right, clean negative space upper left for a headline. |
| What atmosphere? | Light and colour | Soft natural morning light, fresh but realistic greens, warm earth tones, natural skin tones. |
| What should it feel like? | Mood | Positive, practical, community-led, with the volunteers absorbed in the task. |
Day seven starts with a brief
Generating an image has become easy. Writing the brief that makes the image usable is the skill worth having, and it is the part that decides whether your organisation can publish the result.
Today gives you two things: a map of the tools as they stand in September 2026, and one method that carries across almost all of them.
Subject + Style + Framing + Light and colour + Mood + Constraints
We will take one ordinary idea, a community project planting trees in local neighbourhoods, and turn it into a campaign visual you could actually use.
What happens when the brief is one line
You open an image generator and type:
Make me an image about volunteering to plant trees.
You may well get something attractive. You have also handed the model every creative decision in the campaign:
- Who are the volunteers, and what do they look like?
- Are they in a neighbourhood park, on a city street, or somewhere rural?
- Is this photography, illustration or a poster?
- Are people planting, or smiling at the camera?
- Is the campaign optimistic, urgent, community-led or institutional?
- Is there a forest behind them? A globe? A giant green leaf?
Whatever you leave out, the model decides for you. That is the same lesson as day one, in visual form: a gap is an invitation to invent.
The fix is a better brief. Length matters far less than deciding those things in advance.
Every prompt from this bootcamp is collected in the AI prompt library.