Special Events | Summer 2025

That matters; not just ethically, but legally: • In the U.S., consumer data privacy laws (like California’s CCPA) require careful handling of client information. • In the EU, under GDPR, even uploading personal data to a platform without appropriate safeguards could constitute a breach. What you can do: A one-page internal AI policy document is sufficient to get you going. It should define: • What tools are allowed (public vs. enterprise) • What data is off-limits (client info, financials, strategy docs) • What must be reviewed by a human • What AI use must be disclosed Build it into: • Employee onboarding • Vendor contracts • SOPs and project workflows • Quarterly team check-ins You don’t need to be an AI expert to lead; you just need some structure, transparency, and clear boundaries. Creating an AI policy keeps you in control of your brand voice, your client experience, and the standards that set your business apart. It’s not about fear or avoidance. It’s about ensuring your tools, team, and tech align with your values. Christie Osborne is the owner of Mountainside Media, a company that helps event industry professionals and brands develop scalable marketing strategies that bring in more inquiries and leads. Christie is a national educator with recent speaking engagements at NACE Experience, WIPA, ABC Conference, and Catersource + The Special Event. Christie regularly shares industry insight in her Special Events column, as well as on Wed Altered, Rising Tide Society, WeddingIQ, and NACE’s industry blog.

ends up with content that sounds very similar to yours. What you can do: Set a hard rule. Proprietary content and sensitive data do not go into public AI tools without your approval and consent. If you wouldn’t post it on Reddit or drop it in a group chat with your competitors, it doesn’t belong in a prompt box without your oversight and consent. If someone needs inspiration from past work, train them to: • Extract general themes • Summarize frameworks • Paraphrase, don’t paste AI is powerful, but when used without oversight or consent, it can quietly break parts of your business and brand. The efficiency trap: when speed undermines excellence AI-generated content looks polished. But it can also strip away the personalization that modern clients expect. Examples: • A website uses vague copy with no real value proposition. • A templated timeline causes confusion for a unique venue. • A nurture email seems cold and transactional. These moments don’t always set off alarms. But over time, they make your work feel more generic— and slowly chip away at the care, creativity, and quality that used to set you apart. Why this matters: Clients aren’t just hiring you to deliver. They want to feel like their event matters, and when your messaging and communication style starts to sound like it was lifted from a template, it signals that maybe the experience will be, too.

What you can do: Draw a line between what’s efficient and what must remain human-led. Define: • What must be written or reviewed by a human (e.g., nurture sequences, proposals, custom menus) • When AI is allowed for drafting only—not for final voice • How to document or disclose AI use with vendors or clients Examples: • “Our design decks are 100% human-generated.” • “We use AI for outlines only— never for finished copy without full edits.” • “No AI-generated emails to vendors or clients—ever.” This isn’t about rejecting AI. It’s about fostering trust in your brand. The policy gap: what you don’t know can hurt you AI use doesn’t always look risky; it looks like initiative. An assistant pastes sensitive client data into ChatGPT to “analyze trends.” A freelancer reuses your content as a prompt for another client. A contractor builds an AI agent to auto-respond to reviews— without your input or approval. These aren’t malicious moves, but they set precedents you never signed off on. Why it matters: Without clear policies, you can’t track or manage how AI is being used in your business. Even if nothing explodes, inconsistency compounds. Before

long, your voice, your client experience, and your service

standards start to feel disconnected. When sensitive data—like financials, client names, or project history—is pasted into tools like ChatGPT, it’s now in a system you can’t audit, trace, or permanently remove.

9

SPECIAL EVENTS SUMMER 2025

Powered by