Enablement Hub
Learn
LynxAI
Tools
Insights
AI Champions
Governance
Open LynxAI
Request an AI Project
Home
Governance
Frequently asked questions
EU AI Act · Chapter 11 of 11
Frequently asked questions
The most common questions — answered.
Do I need to label AI used for brainstorming or mood boarding?
No. If you use AI solely for internal ideation or brainstorming, you do not need to label the final output. Only final, client-facing deliverables are subject to disclosure.
Is Photoshop "Generative Fill" always Generative AI?
Not necessarily. Standard editing (removing a stray object, background expansion) is Assistive. Changing the meaning of an image (swapping a model, inserting new objects) is Generative.
Who should sign the Human Review Certification?
The person with final editorial responsibility for the delivery — usually the Lead Producer, Medical Writer, or Account Lead. They hold full liability.
Will disclosing AI use undermine the quality of our work?
Absolutely not. Transparency builds trust — clients who know our process trust us more, not less.
Where exactly do I store the audit trail?
Keep all audit trails — Tool Logs, Prompt Summaries, and Certifications — within your project's internal documentation folder. Never in the final client delivery.
Who is liable if AI-generated output contains medical misinformation?
The AI system is never liable. The human reviewer holds full editorial responsibility — treat every AI output as a draft that needs the same rigorous verification as any other work.
Decision matrix & audit requirements
