How to create AI marketing images ethically?

The Short Version

Ethical AI image creation requires three parallel tracks: transparency with your audience, responsible sourcing of training data, and clear internal governance. Many CMOs treat AI images like any other asset, but they carry unique ethical and legal considerations that demand deliberate processes.

Why This Matters for CMOs

AI-generated images are increasingly indistinguishable from real photography. This creates three specific risks:

• Legal exposure: Using AI images that mimic real people without consent can trigger right-of-publicity lawsuits
• Brand trust: Audiences discovering undisclosed AI images feel deceived, damaging credibility
• Regulatory pressure: The FTC and EU are tightening disclosure requirements for AI-generated content

The CMOs winning with AI images aren't hiding the technology—they're being transparent about it and building it into their brand narrative.

Step 1: Establish Your Disclosure Policy

Decide upfront how you'll communicate AI use to your audience:

1. Determine disclosure thresholds — Will you disclose all AI images, or only those that could be mistaken for real photography? (Recommended: disclose all)
2. Choose your disclosure method — Add a watermark, caption, or metadata tag (e.g., "AI-generated image" or "Created with DALL-E 3")
3. Document your policy — Create a one-page internal guide so all team members follow the same standard
4. Test audience reaction — A/B test disclosure statements to see which language resonates without damaging engagement

Example disclosure language: *"This image was created using AI to illustrate [concept]. It does not depict real people or events."*

Step 2: Audit Your AI Tool's Training Data

Not all AI image generators are created equal. Before committing to a tool, verify:

• Training data sources — Does the tool use licensed images, public domain content, or opt-out mechanisms? (DALL-E 3, Midjourney, and Adobe Firefly publish this; others don't)
• Consent and compensation — Are artists whose work trained the model being compensated? (Adobe Firefly compensates creators; others face lawsuits)
• Opt-out availability — Can photographers and artists request their work be removed from training data? (This is increasingly standard)

Recommended tools for ethical sourcing:

• Adobe Firefly — Trained on Adobe Stock and public domain images; compensates creators
• DALL-E 3 — Uses licensed images and has built-in safeguards against generating people who look real
• Midjourney — Transparent about training data; allows artist opt-outs
• Avoid — Unnamed or closed-source generators that won't disclose training sources

Step 3: Establish Guidelines for Human Representation

This is where most ethical issues arise. Create clear rules:

Never generate images that:

• Depict real, identifiable people without explicit consent
• Mimic specific celebrities or public figures
• Create deepfake-style images of real individuals
• Suggest real people endorse products they don't actually use

Safe practices:

• Generate abstract or stylized human figures (illustrated, cartoon, artistic styles)
• Use AI to enhance or modify images of people who've given written consent
• Clearly label any AI-modified images of real people
• When in doubt, use stock photography of real people instead

Step 4: Build Internal Governance

Create a simple approval process:

1. Designate an AI image owner — One person (usually a creative director or marketing ops lead) who reviews all AI-generated images before publication
2. Create a checklist — Does the image have proper disclosure? Does it violate any guidelines? Is the source tool ethical?
3. Document everything — Keep records of which tool created each image, when, and for what purpose (protects you legally)
4. Train your team — 15-minute session on your disclosure policy and tool guidelines

Step 5: Consider Your Brand Narrative

Some of the most successful AI image campaigns lean into the fact that images are AI-generated:

• Transparency as differentiation — Position your brand as honest about AI use
• Artistic merit — Frame AI images as a creative medium, not a shortcut
• Cost transparency — Explain how AI images help you invest more in other areas (e.g., "We use AI for concept images so we can spend more on customer research")

This builds trust rather than eroding it.

Common Pitfalls to Avoid

• Assuming "AI-generated" means "copyright-free" — AI images can still infringe on training data sources
• Forgetting metadata — Remove identifying information from AI images before publishing (prevents reverse-image searches that reveal the tool used)
• Mixing real and AI without disclosure — If your campaign uses both, disclose which is which
• Ignoring cultural sensitivity — AI tools can perpetuate stereotypes; review images for bias before publishing

Tools and Resources

• For disclosure: Watermark tools (Canva, Adobe Express) or caption templates
• For auditing: Check each tool's transparency report or FAQ
• For governance: Create a simple Google Form or Airtable checklist for approvals
• For legal guidance: Consult your legal team on right-of-publicity laws in your jurisdiction

Bottom Line

Ethical AI image creation isn't about avoiding the technology—it's about being intentional with it. Disclose AI use, source from ethical tools, establish clear guidelines for human representation, and document your process. CMOs who build transparency into their AI image strategy build audience trust and reduce legal risk. The brands winning with AI images aren't hiding it; they're being honest about it.