Agentic AI is having its moment.
Every week, another headline or product demo promises artificial intelligence that can research, make decisions, complete the work and hand off the finished result.
In fact, a spring 2025 survey from MIT Sloan Management Review and Boston Consulting Group found that 35% of respondents had adopted AI agents by 2023, with another 44% expressing plans to deploy the technology soon. And agentic AI is expected to create a “multi-trillion-dollar opportunity” for many industries, from medicine to software engineering, according to Jensen Huang, CEO of technology firm Nvidia.
And the travel advisor industry could use the help, right? Especially when the administrative work only seems to grow year after year.
The hype around agentic AI can make it feel much further away, more technical, more out-of-reach and more expensive than it really is. But travel advisors do not need to hire a developer or build a complicated network of digital employees to start using agentic AI.
The reality is, you already have access to agentic capabilities inside the platforms you already use.
What Exactly Is Agentic AI?
Let’s start with a quick primer. When using generative AI, such as ChatGPT or Claude — which, by now, most people are somewhat familiar with — a user gives AI a prompt, and then it responds. For example, you might copy and paste your notes from a client call into Claude and ask it to draft a follow-up email; it creates the email, and the chat ends.
Agentic AI, meanwhile, gets handed a goal and runs with it, taking action with minimal human supervision. It maps out the steps, picks the tools, makes the calls, checks its own work and keeps going without a user babysitting every click.
You Have More Access to Agentic AI Than You Think
Claude Cowork and ChatGPT Work are agentic AI tools that are built to move beyond basic chat features. These assistants were created to take a goal, create a plan, use the tools available and work through the requisite steps.
On a desktop, that might look like the AI opening a browser, navigating websites, working across tabs, reading files and creating documents. For example, instead of inputting a question such as “Can you draft an email featuring this supplier promotion?” to a regular chat window, an advisor could have the agentic AI tool scan their inbox for any new updates from that supplier, assess what is relevant to their ideal client, draft a newsletter written in their voice, and drop that draft into Kit or Flodesk or add it to Asana or ClickUp, assigning it out to your virtual assistant. And it can do this every week, on a schedule.
The Time to Start Experimenting Is Now
As someone who teaches travel pros how to implement AI into their travel businesses, one thing I’ve noticed across all my students and clients as they get started is that they are underusing their platform of choice.
In my opinion, we are in the golden age of AI access and pricing. That $20 per month for ChatGPT or Claude? It’s worth a lot more than 20 bucks. For moderate to power users who really take advantage of all the features available, it’s significantly cheaper than doing the same via API costs.
It’s reminiscent of the early days of Facebook ads. During its initial rollout, I was building my DMC, North & Leisure, and I was in many Facebook groups for business owners, trying to learn how to use ads. I watched as some owners quickly learned how to use them, and because Facebook ads were so new and cheap, those who figured it out were practically printing money.
Then the ads took off. Competition increased, and advertising costs climbed. Today, achieving the same results as the early days requires more skill and a much larger budget.
I suspect we’ll eventually look back at today’s AI subscriptions in a similar way. As these AI companies compete for market share, usage and features are easily accessible, making it the perfect time to experiment and learn.
Getting Started
Begin your agentic AI experimentation with a process you repeat frequently and resent doing manually. It should have a clear starting point, a predictable result and enough repetition to make improving it worthwhile.
Map out what you currently do, including the steps you normally complete without thinking about them. Then, decide which steps require your expertise and which involve sorting, summarizing, formatting, researching or drafting — those are often the easiest places to bring in AI. Next, look at what the AI requires to do the work properly; it may need your templates, brand guidelines, service standards, preferred structure or examples of strong past work. Finally, decide where you need to review the result.
AI is only as useful as the system surrounding it. And while agentic capability means it can do more for you, that does not necessarily mean it's hands-off. Pauses in workflows to allow for human input are to be expected; build in points where you add your distinct POV and approvals/edits.
The bottom line, though? Just get started. Start messy and improve as you go. The major platforms roll out updates constantly, allowing you to do more and stack up the skills and tasks you’ve already trained it to do. For more tips, you can find me on Instagram at @travelprotheory or at my website,TravelProTheory.com.
