AI Use

Romeo is built on artificial intelligence. We think it's important that you understand how the AI works, what it does well, and where you should be cautious. This isn't a legal disclaimer dressed up; it's a real description of what you're using.

1. What's actually happening when you chat with Romeo

When you send a message, your message and your relevant trip preferences are sent to one or more large language models operated by our AI providers. The model reads the input, predicts a helpful response, and sends it back. We then format and display that response in your chat.

Romeo is not a search engine. The model has been trained on a large body of text, but its training data has a cutoff date. It does not browse the live internet for every answer.

2. Where Romeo is reliable

Romeo is generally reliable for:

• Suggesting itineraries based on common destinations and travel styles
• Comparing trip approaches (cities vs beaches, fast vs slow, etc.)
• Recommending broad timing windows (rough seasons, weather patterns)
• Producing structured plans you can use as a starting point
• Remembering your preferences and applying them to future trips

3. Where Romeo can be wrong

Some things the AI is genuinely bad at, and we tell you up front:

• Visa requirements. These change frequently. Always check official government sources before booking flights.
• Entry rules and quarantine. Especially after the pandemic, rules shift. Check official sources.
• Vaccination requirements. Check the CDC, WHO, or your country's equivalent.
• Real-time prices. Romeo can show approximate price ranges but not live fares.
• Specific business hours, current closures, or short-term changes.
• Safety advisories during active conflicts, natural disasters, or political instability. Always check government travel advisories (Smartraveller for Australia).
• Medical and accessibility advice. Talk to a qualified professional.

4. AI hallucinations

Sometimes AI models confidently produce information that's wrong: a restaurant that doesn't exist, a hotel at the wrong address, a tour that closed years ago. We work hard to reduce this, but we can't eliminate it. If something Romeo tells you seems off, double-check before booking.

5. Your data and AI training

We do not use your personal data to train AI models. Our AI providers have committed contractually that they don't train their models on data we send through their commercial APIs.

We may use anonymised, aggregated data to improve Romeo's prompts and routing. This data cannot be linked back to you and isn't used to train models on your personal preferences, trips, or messages.

We may add or change AI providers as the technology evolves. If a change materially affects how your data is processed, we'll update our Privacy Policy and notify you.

6. Liability and AI advice

Romeo's responses are not professional advice. They're a helpful starting point. Before making decisions that depend on accurate information (booking flights, applying for visas, taking medications, choosing insurance), verify with the appropriate official source or professional. We are not liable for losses arising from acting on AI-generated content. See our Terms of Service for the full limitation.

7. Reporting AI mistakes

If Romeo gives you bad information, we want to know. Email hello@withromeo.com with what went wrong. It helps us improve the prompts and routing for everyone.