InEvent探讨活动规划中AI的实际应用：从叙事到现实的专家对话

Posted on October 21, 2026

Imagine a robot helper that never gets tired. It can write, sort, count and plan in a few seconds. Sounds amazing, right? But does it really work that way for the people who plan real events, with real clients, real budgets and real deadlines?

We wanted a straight answer. So at the InEvent Open House on October 8, 2026, called “The AI Narrative vs. AI Reality,” our co-founder Vinni Neris sat down with two people who use AI for event planning in their daily work: Katie McIntyre, Director of Technical Architecture at Opus Agency in Portland, Oregon, and Wendy Porter, Founder and Chief Event Strategist at Wendy Porter Events in Minneapolis.

The short version: AI helps a lot when the data is good and a person stays in charge of the result. That is our view too, and the rest of this article explains why. The narrative says AI will do everything. The reality is quieter, and we think more useful. AI is a strong helper for the jobs nobody enjoys and a poor boss for the jobs that need a human.

Below are the real examples Katie and Wendy shared, plus our take on each. Want the product view too? Read about artificial intelligence in event management at InEvent.

Table of Contents

What is the difference between generative AI and agentic AI?

Two workflows compare generative AI drafting an event plan with agentic AI creating tasks in connected software, both with human review.

Generative AI makes things when you ask. Agentic AI can take a job and do it on its own.

Generative AI responds to your prompts and tries to match what you asked for. It may ask you questions along the way. Agentic AI has some autonomy: you hand over a request, and it decides its own next steps and acts on them without checking in every time. Like Katie said, the clue is in the name. Agentic AI has agency, which means it can act by itself.

Here is a way to picture it. Say you ask a helper for a birthday card. A generative helper asks, “What color should the sun be?” and waits for your answer. An agentic helper hears “make a birthday card,” then picks the colors, draws the picture and hands you the finished card. (That example is ours, not Katie’s.)

Why does this matter? We think it is one of the most useful differences to understand, because it changes what you owe the tool. The more a tool decides by itself, the more you need to check its work. Keep that in mind, because it comes up again below. Want the technical version? IBM has a clear explainer on agentic AI vs. generative AI.

Is AI a threat or an opportunity for event professionals?

It is both, and which one you get depends on who is holding the tool. Wendy leans toward opportunity. Katie answered “yes” to both.

Wendy looks at AI as a business owner. She wants it to take over the manual, time-hungry parts of running a company, like going through her emails each day and listing what needs a follow-up. Less busywork means more time for the parts only she can do.

Katie’s hammer is the picture we keep coming back to. A hammer is very useful to a skilled carpenter. In Katie’s own hands, she joked, it means a quick trip to the emergency room. Same tool, different result, because skill and intention change the outcome. Would you hand a hammer to someone who has never built anything and expect a house? Probably not.

Our take: we land with Wendy on opportunity, with Katie’s warning attached. Vinni made a similar point during the session. Every new technology brings an opportunity, and a threat for anyone who does not respond. Event professionals have already learned plenty of new tools over the past decade, and AI is the next one. We’ve written more about where AI is taking the event industry. Our guide to AI trends in event technology shows what is already out there. The best part of the job is also the part AI cannot do. Events are where the personal side flourishes, and human interaction is not something a machine can replace.

How are event planners actually using AI today?

They use it for the slow, repeat-it-again work, then add their own judgment on top. Here are six real examples from the session. What struck us is how unglamorous they are, and we mean that as a compliment. There is no robot butler here, just hours handed back.

1. Project plans, budgets and summaries

Wendy uses AI to build project plans, draft budgets and write notes and summaries, the manual jobs that eat so much time in events. Here is a tip from the session: AI can take the plan from an older, similar event and update it for today’s. Add the key dates, and it can work backward to fill in the rest.

It is not magic, though. The day before the Open House, Wendy gave AI her own budget template plus a client’s numbers and asked it to fill the template in. It “didn’t actually do as well as I had hoped,” she admitted, and she plans to train the tool more. For now, it is hit or miss.

This was our favorite moment of the whole session. Anyone can show a perfect demo. A pro with years of experience admitting that a test fell short tells you far more about where AI really stands today.

2. Proposals and RFPs

A proposal is the document you write to win a client’s business. Assessing a request and writing one used to take Wendy about 30 hours. Now it takes about five (“let’s say,” she added), a figure she plans to track by loading in timesheets from past events. Even as a rough estimate, that is a big gap. It lets her answer more requests, faster, and get some work-life balance back.

Katie’s team takes a more hands-on route: let AI get you about a quarter of the way there, then bring in people. “Human passion is just about the most attractive thing on the planet,” she says, and she would happily listen to someone talk about their stamp collection for hours if they truly loved it. Proposals work the same way.

So what happens in that first quarter? AI lays the foundation: who the client is, what they have wanted in the past, and how your services line up. Then people add the part that wins the work.

We agree with Katie. In our view, the smart question is not “How much of this can AI write?” It is “Which part should only a human write?” For more on writing proposals, see our event proposal guide.

Ready to build your own? Our event RFP template guide covers the buyer’s side of the process, and our event proposal template gives agencies a structure for responding.

3. Boring setup work with no decisions in it

Katie’s favorite use is event logistics. She used to type a thousand choices into a dropdown menu, one by one. With newer connections called MCPs (they let AI tools work inside other software), she can now hand over a list and say, “please add these.” It is “literally put thing A into place B.”

Why does that excite her? That job needs no human curiosity, because there is no decision in it. We think that makes a great rule of thumb. If a task has no decisions in it, hand it to AI. If it takes judgment, keep it for yourself.

4. Finding information hiding in 50 places

Planning an event means collecting details from everywhere: a prospectus from the show owner, emails from clients, messages from vendors. Wendy counts about 50 places she has to toggle between to pull the puzzle together. She hopes agentic AI can scan all of it, pull out the important nuggets and drop them into one project plan. Think of a tidy-up helper for a very messy desk. We think this may be where agentic AI earns its keep for event teams. It is not flashy. It simply gathers scattered information so people can get on with thinking.

5. Spotting blind spots

Clients with well-organized data get the most out of AI, Katie has found. They use it to challenge their own assumptions and to surface things that “quietly fell off the grid” years ago without anyone noticing. Run the same event year after year, and those gaps get easy to miss.

AI is very good at finding patterns. So her clients also use it to study last year’s attendance, plan schedules, close the gaps between sessions, choose session lengths, and research markets, price points and competitors.

Here is our take: this is the strongest argument for getting your data in order. A tool that finds patterns across years can only work with years of data you can trust.

6. Choosing who gets the VIP invite

Every event with an executive program has the same problem: each salesperson argues that their client deserves a seat. Katie described a fairer approach. Take your CRM data (the records of who your customers are and what they do), make it partly anonymous for privacy, and ask AI who is actually active and showing signs they want to buy. You stop picking the person with the fanciest job title or the most famous company. She called this a “silo buster,” because it gets teams out of their separate corners and looking at the same facts.

We like this one because it swaps office politics for evidence. We would still want a person to make the final guest list.

Where does AI save event planners the most time?

The audience guessed content. The two experts picked logistics and planning. In our live poll, attendees chose content and communications first, then data reporting and follow-up.

Katie and Wendy gave different answers. Katie picked logistics, for the thousand-options reason above. Wendy picked project planning, estimating how long a job will take, and budgets. She also likes AI for brainstorming and for polishing her writing. A math major by training, she does not think of herself as a great writer. But she knows what she wants to say, and AI helps her tighten it so it sounds like her, only crisper.

Surprised? Our guess is that content is the visible work, so it gets the attention, while the unseen jobs quietly give the most time back.

We see the same thing with follow-ups at InEvent. A generic “thanks for coming” email is easy to send and easy to ignore. We have been trying something different with trade show leads. We look at what conversations we had with a lead, and which sessions or dinners they joined. Then AI helps us write a message about what they actually experienced. The goal is a follow-up that feels personal, not copied. For starter templates, see our post-event email guide. Working trade shows? Our guide to tradeshow lead management covers the whole journey, from capturing a lead to nurturing it.

What separates teams that get real value from AI?

Good data, clear intentions and curiosity. That is Katie’s formula, and it is the drum she bangs whenever AI comes up. We think she is right, and the order matters.

Start with the data. “I don’t want to have the AI conversation with you if we haven’t already had the data conversation,” Katie says. It is a case of garbage in, garbage out. Think of baking a cake: start with spoiled eggs, and no oven in the world can save it. Some of her clients feed AI messy information, get messy answers, and decide AI is a scam. “It’s not the tool,” as she puts it. “It’s the wielder.”

Then add the past. AI cannot predict the future by itself. It needs context, like what a client asked for before and what happened at earlier events. We have seen this in our own work. Our team uses AI to spot the questions customers keep asking and the places where our answers fall short.

Next, know what you want. Katie’s clients with the best results know what they are looking for, and then let AI sort the wheat from the chaff. A person still has to taste the wheat and say, “Yes, this is good.”

Finally, play. Curiosity and experimenting help people get comfortable with these tools, Katie agrees, as long as you go in with a fairly clear idea of what you want to find. Our advice: try things, but know your goal.

Which AI tools do event planners use?

There is no single winner, and we think that is the honest answer. Katie’s view is that the best tool depends on the job. Here is what each person uses.

Wendy uses ChatGPT. She started there, and she has so much information built in that she is reluctant to switch. Recently she had it lay out a menu for a client, then moved the result into Canva to tidy it up and make it perfect. ChatGPT’s graphics have improved a lot since last year, she noticed. AI gave her a quick starting point, and she added the human finishing touches.
Katie uses several AI tools and likes to pit them against each other. She will tell one tool what another one said and ask what it thinks. Each has its own personality, she jokes. She also enjoys Adobe Firefly for images.
At InEvent, our team works in Google Workspace, where Gemini helps us make presentations and pull in files from Google Drive and email.

So which tool should you pick? Katie’s answer was honest, and a little funny. She typically doesn’t recommend “a hammer unless I understand the carpenter.” Whether one tool beats another depends on what you need, how much you want to change the finished product afterward, and what you are willing to pay.

Our take: stop asking “Which AI is best?” and start asking “Which job do I need done?” These tools also change fast, as Wendy’s graphics example shows, so today’s favorite may not be next year’s.

How should event teams handle AI, data privacy and transparency?

Be open about it, let people choose what they share, and bring it up early. Would you hand your diary to a stranger without asking first? Attendees feel the same way about their information.

Katie’s approach is transparency and empowerment. Let attendees decide how much of their information your AI gets to see. Plan how to anonymize the data before it arrives, so you can slice it in a way that protects each person’s privacy. And have a real way to act when someone asks to opt out. She also named a fear that goes beyond data leaks: the internet could build a convincing fake version of you from your data, one that is very hard to disprove.

Wendy keeps guardrails in place so her work stays private and does not feed the wider learning models. She anonymizes data wherever she can, so she gets the general patterns she needs without pinning anything to a named person. And she believes these talks belong at the very start of a client relationship.

She was honest about one gap. When Vinni asked whether she tells clients that she uses AI, she admitted those conversations have not happened much yet, and that they need to.

This is not only a small-team question. BizBash’s report from Dreamforce 2026 shows one of the biggest conferences in tech facing the same issue. An executive at Left Field Labs, the experiential agency behind the show floor, said the brands that handle AI controversy best do not gloss over it. They face it head-on.

Our take: put AI and data on the kickoff agenda, right next to budget and timeline. The conversation feels awkward only the first time, and it protects everyone’s trust. If your team has not had it yet, you are in good company, and now is a good time to start.

What should AI not replace in events?

Human review and human connection. Both guests agreed that AI should never run on autopilot, and that face-to-face events are not going anywhere.

Wendy’s story shows why. A client once sent her research with the note, “Here’s what I found.” When Wendy double-checked it, the information was wrong. The AI had pulled very old facts, and the client was about to make a decision based on them. Her rule is simple: “anything it spits out, it needs human oversight and review.” Think of AI as a very fast student. It hands in homework in seconds, but someone still has to check the answers.

Relationships matter too. “I don’t think we’re ever going to get away from face to face events,” Wendy says, and we agree. Human engagement is human engagement, and no tool replaces it. That is why we like tools that help attendees find the right people to meet, such as AI-powered matchmaking. The tool suggests the match. The conversation is still human.

Then Katie asked the question we keep thinking about. Instead of “What can AI do?”, she asks, “What should it do?” She described an idea she had read about: an AI assistant that works like a contextual wallet. It knows what you like, looks through the session list, picks the best sessions for you and even books them. That sounds handy, like another door into the event. But as Katie wondered, “At what point does it stop being a door and start being a gatekeeper at the door?”

Imagine your AI decides you do not like sponsors, so it never shows you anything from them. What do the sponsors and exhibitors get for their money? Our take: this is one of the biggest open questions in event technology, and anyone building these tools should take it seriously. Katie admitted she does not have a final answer yet. For how personalization works when attendees give their authorization, see our post on AI and personalized attendee experiences.

She also made a case for keeping a little friction. Friction is the bumpy, slow part of learning, like practicing a skill until it sticks. “Friction is how we learn,” she says. As a web developer, she does not memorize every detail, because she knows where to look things up. But handing off everything makes the experience less human, less valuable and less memorable.

We think she is onto something. Remove the friction from chores, like typing a thousand dropdown options. Keep it where learning and connection happen, like conversations and discovery.

How can event planners start using AI?

Start small, start with clean data, and keep a person in charge. If you are wondering how to use AI for event planning, here are six steps drawn straight from what Katie, Wendy and Vinni shared. If you only do one, do the first.

Have the data conversation first. Before you try AI, check that your information is organized and trustworthy. Good ingredients make a good cake.
Begin with boring, no-decision jobs. Importing lists, updating an old project plan and writing summaries are good places to start.
Let AI get you part of the way. Then add your own voice, your passion and what makes your team different.
Check everything it gives you, especially research. Old or wrong facts can look very convincing.
Agree on privacy and AI-use rules with clients at the start. Anonymize data where you can, and give attendees a say in what they share.
Stay curious and match the tool to the job. Try things out, but know what you are trying to get done first.

Frequently asked questions

1. How can you use AI for event planning?

According to the Open House guests, planners use AI for project plans, budgets, summaries, first drafts of proposals, importing lists and pulling scattered details into one plan. They suggest starting with repetitive jobs that need no decisions, then reviewing everything the AI produces before it goes out.

2. Will AI replace event planners?

The guests did not predict that. Wendy called AI an opportunity, while Katie said it is both a threat and an opportunity, depending on the skill and intention of the person using it. Both said AI output needs human review, and Wendy said face-to-face connection cannot be replaced.

3. What is the difference between generative AI and agentic AI?

Katie explained that generative AI responds to your prompts and tries to match what you ask for. Agentic AI has some autonomy: it takes a request and acts on its own, without checking in at every step. Because of that, agentic AI needs careful review of its work.

4. What tools do event planners use for AI?

Wendy uses ChatGPT for writing, ideas and first drafts, then finishes designs in Canva. Katie uses several tools, including Adobe Firefly for images. Vinni said InEvent uses Gemini through Google Workspace. Katie stressed that the best tool depends on the use case.

5. What data do you need before using AI in events?

Katie said to have the data conversation before the AI conversation, because poor data produces poor results. Clients with mature, well-organized data use AI to find gaps in their assumptions and patterns across years. Both guests also recommended anonymizing attendee information where possible.

Join us at the next Open House

The best part of this session was the honest back-and-forth, and we want more of it. Join us at the next InEvent Open House to hear event leaders talk about what is working, what is not, and what is still a big question mark. And if you are curious how AI fits into event software, explore AI in event management at InEvent.