Aurentex. We build the tech. You run the event.
How it worksPricingOur storyBlog
Operations

AI in event operations: the unglamorous jobs where it actually pays

Nearly every event team expects to use more AI this year. Far fewer say it has changed much yet. The gap is not about which model you pick. It is about which jobs you give it, and what state your data is in when you do.

The Aurentex Team
28 September 2026, 6 min read
Delegates in business suits seated facing a blue-lit stage screen at a corporate summit, some working on laptops
The useful AI at an event is mostly invisible to the people in the room.

Every 2026 trend piece on events leads with AI, usually with a picture of a chatbot greeting attendees. The event teams actually using it describe something duller and more useful: a faster first draft of the reminder email, a quicker answer to "who registered but never showed up", a post-event report assembled in an afternoon instead of a week.

What the surveys actually say

Adoption is no longer the question. The Northstar Meetings Group and Cvent PULSE survey found 65% of planners using generative AI tools.1 Bizzabo, an event platform vendor, reports that 95% of organisers in its 2026 survey expect their use of AI in events to increase.2 American Express Global Business Travel's forecast, based on a YouGov survey of 601 meeting professionals, puts the share integrating AI into planning and execution at about half.3

Results are another matter. In the same PULSE survey, only 16% said AI had significantly improved planning and execution, while 49% reported modest gains.1 The ICE benchmarking report on corporate event teams found 44% using AI moderately and 31% still piloting.4 Bizzabo notes concerns about accuracy, transparency and data governance are growing alongside adoption.2

65%
of planners are using generative AI tools1
16%
say AI has significantly improved planning and execution1
31%
of corporate event teams are still piloting AI4

Put together: most teams have tried it, and only a small share say it has changed much. That is not a verdict on AI. It is a sign that the easy use, typing a question into a chat window, runs out quickly.

Three jobs where it pays

The pattern in the jobs that work is the same: a person checks the output before it reaches anyone, and the task is one where a good first draft saves most of the time.

Attendee communicationsConfirmation, reminder, venue-change and follow-up emails, in the register of your brand, and rewritten for the segment that is receiving them. Communications is one of the two most-cited expected uses in the Amex GBT survey, at 35%.3 Someone still reads every email before it sends.
Questions about your data"Which companies sent three or more people?" "Who registered for the workshop and did not check in?" "Which email did the no-shows never open?" Questions like these used to need someone comfortable with pivot tables. They are the strongest case for AI in operations, on one condition covered below.
Post-event reportingTurning attendance, session scans, email results and feedback comments into the report your leadership actually reads. AI is good at summarising three hundred free-text feedback answers into themes. It should not be trusted to calculate the headline numbers; those come from the system, and the AI writes around them.
The useful question is not "what can AI do at my event?" It is "which job do I already do every time, where a draft would save me an hour?"

Where it does not

Answering attendees unsupervised. An assistant that tells a delegate the keynote is in Hall B, from an agenda that moved on Tuesday, costs more goodwill than it saves. Attendee-facing answers need a source of truth that updates, and a person who owns it.

Decisions with consequences. Who gets access to the VIP lounge, whose refund is approved, which registration is a duplicate to delete. Let AI flag; let a person decide.

Inventing the numbers. Language models write fluent sentences around figures whether or not the figures are right. Any number in a report should be traceable to a record, not to a paragraph a model produced.

The data decides

Cvent's write-up of the PULSE survey reaches the same conclusion from the other direction: real value comes when AI is paired with clean data, repeatable processes and clear metrics for time and cost savings.1

That is the quiet reason for the modest results. "Who registered but did not attend" is a trivial question if registration and check-in live in one record. It is an afternoon of matching names across two exports if they do not, and an AI given two mismatched spreadsheets will guess. We described that failure mode in running registration on a spreadsheet, and the seams between tools in the 2026 event tech stack.

There is also a data protection question. Pasting an attendee export into a consumer chat tool is sharing personal data with another company, and under the DPDP Act you remain answerable for it. Use business plans with clear data terms, share only the fields the task needs, and prefer tools that query the data where it lives. Our DPDP guide for organisers covers the obligations.

When you do not need it

If you run two or three events a year with a few hundred guests, a good email template, a checklist and one person who knows the event will beat any AI workflow on time and on accuracy. The effort of setting up, checking and correcting AI output only pays back when the same job repeats often enough.

And if your registrations, check-ins and email results live in different places, fix that first. AI layered over scattered data gives you confident answers to the wrong question.

Where Aurentex sits

Today, Aurentex can keep registration, check-in, email tracking and analytics in one record per attendee, on your domain. That is the precondition for every useful AI job above, and it is the part that cannot be bolted on afterwards.

Connecting your own Claude or ChatGPT account to that event data, so you can ask questions of it directly, is marked coming soon on our pricing page, and it is not built yet. We would rather say that here than let a feature list imply otherwise. Your first event is free, which is a reasonable way to see whether the data comes out clean enough to be worth asking questions of.

Common questions

How many event planners are using AI?

Most, for some tasks: 65% of planners use generative AI tools,1 and 95% of organisers in Bizzabo's survey expect their use to increase.2 Only 16% say it has significantly improved planning and execution.1

What is AI actually useful for?

Drafting attendee communications, answering questions about registration and attendance data, and summarising feedback into a report, with a person checking the output each time.

Why are the gains modest for most teams?

Mostly the data underneath. AI can only reason about records it can see, and many events keep them in separate tools.

Is it safe to paste attendee lists into a chatbot?

Treat it as sharing personal data with another company. Use a business plan with clear data terms and share only what the task needs.

Want your event data in one place first?

Registration, check-in, email tracking and analytics in one record per attendee, on your own domain and branding. Your first event is free.

Talk to us
AIOperationsEvent data

Get the next one in your inbox

Occasional, practical notes on event tech. No spam.

Unsubscribe anytime. We will never share your email.

References

  1. Cvent, "What's Trending: Taking the Pulse of the Meetings Industry," 4 February 2026, reporting the Northstar Meetings Group and Cvent PULSE survey. "65% are using generative tools", "only 16% say AI has significantly improved planning and execution, and nearly half (49%) are seeing just modest gains", and the conclusion that real value comes when AI is paired with clean data, repeatable processes and clear metrics for time and cost savings.
  2. Bizzabo, "Event Industry Trends 2026: What Event Professionals Are Reporting," 18 February 2026, summarising Bizzabo's 2026 State of Events Benchmark Report. 95% of surveyed organisers expect their organisation's use of AI in events to increase; concerns around accuracy, transparency and data governance are growing. Bizzabo sells event management software.
  3. Business Travel News, "Amex GBT Meetings Forecast Shows Rising Optimism, Costs," 7 October 2025, reporting American Express Global Business Travel's 2026 Global Meetings and Events Forecast, a YouGov survey of 601 meeting professionals in eight countries. About half are integrating AI into the planning and execution of meetings; event communications and attendee and sponsor matchmaking are each cited by 35% as expected use cases for 2026.
  4. CIM, "ICE Report 2025: Event Organisers Drive Impact," reporting the ICE Annual Benchmarking Report 2025, a survey of in-house corporate event planners supported by Cvent. "AI adoption is emerging (44% using it moderately; 31% piloting)."

The surveys cited use different samples, regions and questions, so their percentages should not be added together or compared as if they measured the same thing. This article is operational guidance, not legal advice.