AI is creating huge opportunities for agencies. It is also creating real legal risks. As AI becomes more deeply integrated into agency operations and client work, questions around intellectual property, data privacy and regulation are becoming harder to ignore: Who owns AI-generated content? What happens when training data contains copyrighted material? How should client data be handled when AI tools are part of the workflow?
Most agencies do not have clear answers yet — and the legal system is only beginning to supply them. Rules are still catching up with the technology, which means agencies are making consequential decisions every day without settled legal guidance.
The questions every agency now faces
Ownership of AI-generated work. The U.S. Copyright Office has repeatedly affirmed that copyright protection requires human authorship: purely machine-generated material is not protectable, while works combining human creativity with AI assistance are assessed case by case. For an agency whose business model rests on selling creative work product, that distinction directly affects what can be licensed or assigned to clients — guidance is laid out in the Copyright Office’s AI initiative.
Training data and infringement exposure. Ongoing litigation over whether training AI models on copyrighted material constitutes infringement remains unresolved, and outcomes will shape which tools agencies can safely rely on and what indemnification they should demand from vendors.
Client data privacy. When AI tools touch customer information, agencies inherit obligations under privacy laws such as the GDPR and U.S. state privacy statutes. Feeding client data into consumer-grade AI tools without contractual and technical safeguards is one of the most common — and avoidable — exposures.
Regulatory change. The EU AI Act is being phased in through 2026-2027, and U.S. state legislatures continue to pass AI-specific rules. Agencies serving clients across borders will feel these requirements first through client contracts and procurement questionnaires.
Expert guidance: the Toerek Law session
These issues were the focus of “AI, IP, and Risk: What Agency Leaders Need to Know Now,” a session delivered by Samantha Jorden, senior associate at Toerek Law — a Cleveland-based firm advising agencies nationally on intellectual property and marketing law — at the AI for Agencies Summit 2026, a free virtual event held on February 12, 2026 and now available on demand from the Marketing AI Institute.
The session covered the intellectual property questions agencies need to understand, data privacy considerations when AI tools interact with customer information, emerging rules and what they may mean for agency operations, and practical steps to reduce legal risk without slowing AI adoption. Jorden’s approach is practical rather than fear-based: deliberate, research-driven exploration of AI, guided by ethical guidelines and documented internal policy.
Practical steps agencies can take now
Put an AI use policy in writing. Which tools are approved, what data may and may not be entered into them, and who reviews AI-assisted output before it ships. A written policy is also increasingly requested in client procurement reviews — a discipline closely related to AI project governance.
Update client contracts. Disclose AI use where material, clarify ownership and warranty terms for AI-assisted deliverables, and review vendor terms for indemnification and data-handling commitments.
Keep humans in the loop on creative output. Human creative direction and editing strengthen both quality and the case for copyright protection of the final work.
Document provenance. Recording which tools and prompts produced which assets makes it far easier to answer client and legal questions later.
Limitations and what to watch
This area of law is moving quickly: court decisions on training data, new Copyright Office guidance, and the staged application of the EU AI Act could each change the practical advice within months. Nothing here is legal advice — agencies with material AI exposure should consult counsel familiar with both marketing law and the fast-changing AI landscape.
The bottom line
AI legal risk for agencies is real but manageable. The agencies that fare best will not be the ones that avoid AI, but the ones that adopt it deliberately — with written policies, updated contracts and human review — while the legal landscape settles around them.
