Sustainable marketing

Artificial intelligence is reshaping how we plan, promote and run marketing strategy. But with great capability comes responsibility. This blog post is a companion to my recent webinar presentation and gives you a practical overview of how to use AI in marketing — efficiently, ethically and in compliance with EU regulations.

What is Responsible AI – and Why Should Marketers Care?

Responsible AI is defined as a set of principles guiding the design, development, deployment and use of artificial intelligence, with the goal of building trust while maximising positive outcomes and mitigating risks.

For marketers, this is not abstract philosophy. Every ChatGPT prompt, every AI-generated visual, every automated email sequence carries an environmental, legal and ethical footprint.

Key environmental facts (source: MIT News, January 2025):

• Each ChatGPT query uses ×5 more electricity than a Google search.
• Training a GenAI cluster requires ×8 more energy than typical computing.
• Data centres consume approximately 2 litres of water per kWh to stay cool.
• New models are released every few weeks — energy spent on prior versions is largely wasted.
• Users currently have no visibility into their personal AI carbon footprint.

The AI Regulatory Landscape: What You Need to Know

Two frameworks are central for anyone using AI in marketing across Europe: the EU AI Act and the GDPR. Together they create clear lines between what is permitted, what requires consent and what is strictly prohibited.

Prohibited practices (EU AI Act, Article 5)

Source: artificialintelligenceact.eu/article/5/

• Manipulating attendee decisions or exploiting psychological vulnerabilities.
• Inferring emotions or mental states (e.g. concluding ‘the candidate is happy’ from facial expressions).

• Biometric categorisation of participants.
• Scraping facial images from CCTV or the internet; deploying deepfakes to deceive people.

Data protection obligations (GDPR & CNIL)

Source: CNIL AI & GDPR Recommendations

• Inform attendees when their data is used to train AI models.
• Uphold rights to access, rectify, object and delete personal data.
• Build privacy protection into your AI tools from the design stage.
• Anonymise AI models wherever possible.

Your action plan

Resources: CNIL AI Checklist  ·  EDPB Guidelines

• Audit all AI tools used for events against the prohibited practices list.
• Add AI data usage disclosure to attendee registration forms.
• Consult CNIL’s AI checklist before deploying any AI solution.
• Follow EDPB best practices for cross-border events in the EU.

Understanding Your AI Options

Not all AI tools are created equal — and choosing responsibly starts with understanding the three main categories available to event marketers today.

• LLM Models — ChatGPT, Claude, Perplexity, Mistral — large language models you prompt directly for content, research and ideation.
• Tools with AI Modules — Salesforce, HubSpot, Canva, Adobe, Notion — AI features embedded inside platforms you already use.
• Agentic AI (the 2026 trend, source: Google Cloud AI Agent Trends 2026) — autonomous agents that connect your tools, complete multi-step tasks and report back with minimal human intervention.

Agentic AI is particularly exciting for event teams: imagine a Content Agent that drafts your social posts from a strategic brief, a Creative Agent that generates visuals, and a Reporting Agent that delivers weekly campaign insights every Friday — all with your oversight.

The 80/20 Rule: Choosing the Right Tool

The AI market is saturated. Before adopting any new tool, ask yourself: does it deliver high productivity for a reasonable cost? Use a simple Cost vs. Productivity matrix to guide your decisions.

• High productivity + low cost: paid LLM plans offering unlimited prompts and full feature access — the sweet spot for most event teams.
• High productivity + high cost: agentic AI setups connecting multiple tools — powerful, but justify the investment before committing.
• Low productivity + low cost: free LLM tiers — useful for quick tasks but require significant human intervention.
• Low productivity + high cost: advanced platforms (e.g. Adobe suite) used without proper training — avoid until the team is ready.

3 Things to Consider for Responsible AI Use

1. Pick High-Impact, Cost-Efficient Tools

Audit your current stack for overlapping solutions. Duplication drives cost without adding value. Reassess tools after a test period: does this accelerate growth, or just add another thing to manage?

2. Monitor for Bias & Stereotypes

AI models can reproduce and amplify gender, racial and cultural biases. Audit AI-generated visuals, copy and targeting for stereotypes. Diversify your prompts and input examples. Assign a human reviewer for all external-facing AI content. Document incidents and refine your guidelines.

3. Protect IP in AI-Generated Content

IP law around AI-generated images is still evolving. Practical steps:

• Use tools trained on licensed or public domain data: Adobe Firefly, Getty AI
• Avoid prompting for named artists’ styles, trademarked characters or logos.
• Run a reverse image search (Google Images, TinEye) on all outputs before publishing.
• Review platform T&Cs — commercial use rights vary widely across tools.

How Leading Companies Are Doing It

Responsible AI is not just for regulators and academics — major corporations are already embedding it into their operations.

• H&M uses a Checklist for Responsible AI on every AI project, helping teams identify potential risks and ensure alignment with company values.
• Allianz commits to avoiding direct and indirect discrimination (on the basis of gender, race and more) and actively mitigates bias in both training data and individual models.
• BMW Group restricts data collection to what is essential for business purposes, and has committed to using AI to reduce emissions and optimise energy consumption.

Conclusion: Use AI Smarter, Not Just More

AI offers event marketers extraordinary capabilities — but the default setting should be thoughtful, not automatic. Before deploying any AI solution, ask yourself: Is this compliant with EU law? Does it respect my attendees’ data? Does it deliver real productivity gains? Is there a risk of bias or IP infringement?

Responsible AI is not a constraint on creativity. It is the foundation for building lasting trust with your audiences — and that is what great event marketing is ultimately about.

About the Author

Connect on LinkedIn: linkedin.com/in/polina-marishicheva