Meeting Highlights: AI &
LAW - MINEFIELD OR NEW OPPORTUNITIES? 22.09.2025
This Cre-AI-tivity session brought together experts from art, law, advertising, and production to address the legal complexities of AI-generated content. Moderated by Marcin Rossa (creative director, GenAI designer and Founder at Creait.me), the panel featured Kobas Laksa (visual artist, director and Golden Lion winner), Magdalena Lissowska-Wołowiec (COO at Creait.me), Krzysztof Słupiński (legal advisor), and Łukasz Gumowski (member of the board at Play Agency and SAR Association of Advertising Agencies AI working group leader).
The key takeaway: while the industry operates in a legal gray zone, transparency, documentation, and internal procedures offer the best protection for creative professionals and their clients.
Key Trends Discussed:
- The Copyright Reality
AI-generated content typically does not qualify as copyrightable work due to lack of direct human authorship. Attorney Krzysztof Słupiński cited a Prague court case where a designer lost after generating an image with only a prompt – it wasn’t considered a protected work.
Simply typing prompts doesn’t create copyright. Only substantial human creative input beyond AI output can potentially qualify for protection. Most AI-generated content exists in an “ownerless domain” – meaning creators cannot transfer “100% copyright” to clients, creating a fundamental contract mismatch with traditional advertising agreements.
- Documentation is Essential
The panel emphasized documenting the entire creative process: all prompts, manual refinements, client approvals, and workflow integration with traditional production. While this may not guarantee copyright protection, it demonstrates transparency and proves genuine human creative involvement – critical for legal protection and client trust.
- Operating in the “Wild West”
The consensus: we’re in a legal gray zone. Current laws were written for a world that no longer exists. Major challenges include:
- No insurance coverage for AI-related IP risks
- Contract requirements (100% copyright transfer) that can’t be met
- Undefined liability when AI content causes problems
- The Real Risk
Lower concern: Someone copying your AI-generated ad.
Critical concern: Unintentionally delivering plagiarized content because AI reproduced elements from training data. As one participant noted: “The biggest risk isn’t someone copying our work – it’s that we deliver work that could be plagiarism and we don’t even know it.”
- Style vs. Character
Legal: Creating content “in the style of” any artist (Ghibli, Wes Anderson). Style itself isn’t protected by copyright.
Illegal: Generating specific copyrighted characters (Darth Vader, Mickey Mouse), even if slightly altered.
Challenges and Opportunities
Key Challenges:
- Corporate policy vacuum: Many corporations lack clear AI policies; contract updates can take over a year
- Knowledge gaps: Rapid development means many professionals don’t understand implications
- Free tool risks: Free AI versions often use uploads for model training, creating rights issues
- Contract chain breakdown: When AI is involved, the entire rights-transfer chain becomes legally misaligned
Major Opportunities:
- Democratization: Professional-grade tools now accessible for 99 PLN/month
- Efficiency gains: AI processed 3,500 product photos per designer per day in one case study
- Pioneer advantage: Early adopters establishing procedures will gain competitive edge
- Hybrid workflows: Combining traditional skills with AI creates results impossible with either alone
Practical Guidelines
Three-Tier Risk Management:
- Professional Tools
- Use paid, enterprise-level services with proper licensing
- Avoid free versions that train on your uploads
- Budget several thousand PLN monthly for professional use
- Internal Procedures Establish clear protocols for:
- Which clients permit AI and under what conditions
- Required documentation standards
- How to present AI concepts
- When AI can be used for final deliverables
- Transparent Client Relations
- Never hide AI usage
- Explain what rights can/cannot be transferred
- Help clients understand risk-benefit balance
- Propose contract amendments reflecting AI workflows
Protection Strategy
Document everything: prompts, refinements, approvals, and full workflow. This demonstrates human input and protects all parties.
As attorney Słupiński noted: “Just because something isn’t a protected work doesn’t mean it violates someone else’s rights. The real risk is unintentional plagiarism by the AI model itself.”
Looking Forward
Regulation Coming: EU AI Act (2026) will likely require labeling of AI-generated content with digital identification.
Market Polarization:
- Professional segment: Enterprise tools, custom models, documented workflows, creative control
- Mass-production segment: Free tools, high-volume content, minimal human intervention
Human Value Remains Central: Despite AI sophistication, human involvement remains essential for creative quality and emotional resonance. Many AI-generated pieces lack soul without human understanding behind them.
Conclusion
The industry operates in legal gray areas while establishing best practices. The safest path forward requires:
- Education – understanding AI and applicable legal frameworks
- Professionalization – using enterprise-grade tools with proper licensing
- Documentation – creating clear records of creative processes
- Transparency – never hiding AI use from stakeholders
- Procedures – establishing workflows accounting for AI-specific risks
As Kobas Laksa observed: “We’re pioneers of a new Wild West – and we’ll be the ones building the towns”. The question isn’t whether to use AI, but how to do so responsibly while maintaining human creativity, ethical standards, and legal awareness.
Partner of this episode: AiAcademy.plus
