AI is dramatically better when grounded in your own knowledge instead of “the internet”

Most AI trip planners are trying to guess what you want. The normal flow is: tell the AI you are going to Japan for two weeks, answer a few questions, let it search the web, and hope the itinerary feels right. And that is useful. Sometimes.

But it is still inference.

AI is much better when it is grounded in your own knowledge, not just the open internet. Think about the results you get when connected to personal knowledge systems like Notion, Obsidian, Granola or Mem versus just random generative answers.

Reely.travel is the first to apply that idea to travel. It starts with the travel inspiration you have already saved from the place you window shop your travels - social media. The Reel you liked. The TikTok that made you want to go. The YouTube video that showed the hotel, restaurant, beach, hike, bar, or neighbourhood that actually caught your attention.

That saved content is not just a bookmark. It is traveller intent.

The flow looks more like this:

Instagram Reel → AI extracts the travel products → maps everything → traveller curates Y/N → trip knowledge base → external AI agents can reason over it. This last bit matters.

With MCP, Reely becomes more than an app. It becomes a traveller-controlled context layer. Instead of asking Claude to “plan my Greece trip” from scratch and having the wisdom of the crowds that the AI is trained on, you instead get specifics like the hotel in Thessaloniki from the reel you saved 4 months ago.

The AI is no longer guessing. It is reading the traveller’s own curated corpus.

This content is provided by the newsletter sponsor reely.travel - you can and should try it for free for your own next trip

“Travel distribution is no longer defined by a single entry point.”

If pumping resources into SEO like it’s 2019 is your strategy then you probably won’t be around in 2029.

For years, most travel businesses thought about distribution in fairly neat buckets. Google. OTAs. Direct. Email. Maybe paid social if someone in marketing had budget left over.

But the traveller does not move like that anymore.

Damien Pfirsch from Agoda points to American Express research showing 75% of travellers say social media influenced their destination choice. That is not a soft brand awareness stat. That is demand being shaped before the traveller ever gets to Google, your website, an OTA, or a booking flow.

And this is the bit I keep banging on about: if social is where intent is being created, then showing up there is not optional. It is distribution.

And you simply can’t get there from just posting on your brand account. You require other voices posting ABOUT you.

Wyndham is selling AI on owner economics - it is framed around owner margins.

The Wyndham angle in the latest digest is not really "look at our cool AI". It is much more practical than that. Lower costs, protect margins, improve the economics for owners. That is a very different conversation from most of the AI puff floating around travel.

The AI projects that actually stick are usually the ones that either make more money, save time or reduce cost in a way the operator can feel. If Wyndham can make AI part of the owner value proposition, not just the guest marketing story, that is stronger than most of the consumer-facing launches we keep seeing. They are selling value to those who hold the chequebook.

The other thing I like here is the implied shift in where the marketplace battle sits. Not just who gets the traveler. Who can give the supply side better economics while still keeping the customer relationship. That is the harder game.

Is Jetstar getting ready for A2A?

Jetstar CEO says charging extra for cabin baggage is about choice and about keeping airfares low. She said this without even a hint of a smirk. That’s impressive in itself.

Never mind most people just choose them because they are cheap. And this makes flying more expensive (but headline price less expensive - hello search engines!)

But there is another angle I keep coming back to.

In an agent-to-agent (A2A) travel world, airlines need negotiable variables.

If your fare is just one bundled price, then an AI agent comparing options has a pretty simple job: find the cheapest acceptable flight. That is brutal for airlines, especially low-cost carriers whose whole model is built around appearing cheap at the first point of comparison.

Unbundling changes the game.

Now the airline is not selling one thing. It is selling a base seat, cabin bag, checked bag, seat choice, boarding position, flexibility, food, maybe bundles, maybe loyalty perks. Each becomes a lever an agent can negotiate, optimise, trade off or package differently depending on the traveller.

That is a very different conversation from “sort by lowest fare”.

So yes, this may just be a money grab.

But it may also be airlines getting structurally ready for a world where machines compare, bundle and negotiate travel on our behalf. In that world, the airline with more components to price has more ways to win than just being cheapest.

Terrible for anyone who just wants the advertised fare to mean the actual fare.

Potentially very useful for the agents

EU introduces new rules on AI - what it means for travel companies

Pay attention to the EU’s new AI Transparency laws - effective immediately. Called the cookie moment for AI - here are some tips on what you need to do:

The 4 transparency rules (in plain English)

The EU AI Act’s Article 50 is essentially about being honest when AI is involved.

1. If people are talking to AI, tell them.
  Chatbots, voice bots and AI assistants must clearly disclose they’re AI at the start of the interaction.
2. If you’re building AI that generates content, it must identify its output as AI-generated.
  This is mainly for AI providers (e.g. OpenAI, Anthropic) who need machine-readable watermarking/metadata.
3. If AI is analysing people’s emotions or biometrics, tell them.
  Applies to things like AI detecting age, gender, mood, stress or facial characteristics.
4. If AI creates realistic fake media (deepfakes) or AI-written public-interest news, disclose it.
  People should know they’re looking at AI-generated or AI-manipulated content.

The takeaway for a tour operator

For 95% of small operators, compliance is actually quite straightforward:

* ✅ Put a clear “I’m an AI assistant” message on your chatbot.
* ✅ Label AI-generated videos or realistic synthetic people if you use them.
* ✅ Human-review marketing copy before publishing.
* ✅ Don’t use emotion-recognition AI without informing customers.

If you’re simply using ChatGPT to help your marketing team write blog posts, social posts, emails and itineraries, Article 50 creates very little additional burden. The biggest practical requirement is making customer-facing AI obvious and not trying to pass AI-generated media off as real.

If you think someone (or everyone) you know or work with could grow from being more informed on the topic of ai + travel (or could use the training above) then please forward this email to them and they can click the button below:

Layla gets snapped up by Expedia

Layla was built around the idea that travel planning should feel like a conversation. Expedia already has the inventory, the booking rails, the customer base, the supply relationships and a mountain of first party data. Put those things together and you get much closer to the full loop: inspire me, plan with me, book it for me, then support me when something changes.

That is the battleground now.

Not who can build the cutest AI interface. Who can connect planning, supply, context and transaction in one place. Expedia have been trying to build trip planning inside their eco-system so interesting that they've now moved to acquisition to fulfil that need.

It also means that choice has gone for Layla users. It is now Expedia inventory and pricing, not choice or negotiation.

Also worth noting what this says about AI startups in travel. If you can build a genuinely good planning experience, there is a pretty clear exit path: the big platforms want the capability, but they want it plugged into their marketplace, not floating around independently forever.

Congrats to Saad Saeed, Sardar Bali and the whole Layla team!

Slack Group!

Podcasts and Sponsors

Podcasts now on Spotify and Apple Podcasts:

All podcasts are now showing up on Spotify and Apple Podcasts for your easy listening pleasure! Now is a great time to binge those you may have missed.

Veerle Witte has done something extraordinary in travel. She has built a successful B2C product. More than 100,000 people are journaling their travel experience - and they might just be doing that whilst travelling with you!

Veerle was the guest on the podcast this week.

Partner with Us - sponsorship info for the newsletter and podcast can be found here.

Most clicked last week was the link to the AMEX GBT story about their MCP server.

That’s it - you’ve made it to the end of this edition. If you’re thankful for this newsletter - you can always buy me a coffee.

I’ll be putting the result of the most clicked post in next week’s edition so you can see where others are focusing. If I’ve missed something, you’ve got a tip or any feedback at all - you can simply reply to this email and it will come straight to me. I’m doing this for You so please don’t be shy to tell me what you think