Agents at the Gate: How AI Is Rewriting the Rules of Airline Fraud Prevention

The airline and travel industry has always been a prime target for fraudsters. High-value transactions, complex loyalty programs, and global customer bases create a rich environment for bad actors. But something fundamental has shifted. The threat is no longer just sophisticated; it's fuelled by AI agents that can adapt faster than defenses.

Now Boarding: The Agentic Traveler

Airlines and OTAs are expecting non-human traffic across the customer journey

• 33% Estimate that between a quarter and a half of their traffic is non-human today.
  10pp above average.
• 61% Expect non-human traffic to increase dramatically or significantly over the next 12 months.
  8pp above average.
• 63% Expect agents to perform account actions such as booking changes/cancellations.
  8pp above average.
  

The New Reality: AI Fraud Is Already Here

For too long, AI-facilitated fraud has been treated as a problem for the future. It isn't. According to Darwinium's AI Survey (February 2026), 97% of organizations have already seen AI-facilitated attacks increase year-on-year, and 75% estimate that more than a quarter of their current fraud attempts are AI-assisted. Perhaps most strikingly, 93% have already encountered deepfake fraud, with more than half experiencing it multiple times.

These aren't edge cases or isolated incidents. They are the new baseline.

The Fraud Stakes Have Changed

AI-facilitated fraud is not a future problem, it’s a reality

• 97% of organizations have seen AI-facilitated attacks increase YOY
• 75% estimate that more than 25% of their current fraud attempts are AI-assisted
• 93% have already encountered deepfake fraud, with more than half experiencing it multiple times

Airlines Are Ground Zero

While AI-driven fraud is surging across every industry, travel and aviation are bearing a disproportionate share of the impact. 17% of airline and travel leaders report that AI-assisted fraud has increased dramatically in the past year, more than double the cross-industry average of 7%.

The reasons are structural. Airlines handle high-volume, high-value transactions. Loyalty point systems represent significant financial assets. Booking flows are complex, with multiple touchpoints across account creation, login, search, checkout, and payment, each one a potential attack surface. Fraudsters have noticed, and they've come equipped.

The Agentic Traveler Changes Everything

There's a second transformation underway that intersects directly with the fraud problem: the rise of the agentic customer. AI agents are increasingly acting on behalf of travellers, searching for flights, comparing prices, making bookings, and managing changes. Airlines and OTAs already know this is happening.

• 33% estimate that between a quarter and a half of their traffic is already non-human. Ten percentage points above the cross-industry average.
• 61% expect non-human traffic to increase dramatically or significantly over the next 12 months. Eight points above average.
• 63% expect agents to perform account actions such as booking modifications and cancellations, again, eight points above the norm.

This is the shape of the near future: a customer base increasingly represented not by humans at keyboards or on phones, but by AI systems acting on their behalf.

The Agentic Paradox

Here lies the central challenge. The same technology that is empowering legitimate customers is empowering fraudsters. This is not a coincidence, it is an intrinsic feature of powerful, general-purpose AI.

Trusted customers will increasingly use agents to make travel decisions, compare products and prices, and complete payments seamlessly. Meanwhile, fraudsters are using the very same capabilities to find weak spots in booking flows, bypass account and login protocols, and generate convincing deepfakes, scam attempts, and phishing attacks at scale.

The problem is compounded by a dangerous knowledge gap. 46% of airlines and OTAs say they cannot distinguish good agentic traffic from bad. And 56% say they cannot authenticate an agent's intent. When organisations can't tell friend from foe in nearly half of their automated interactions, the exposure is immense.

From Transaction-Level Thinking to Journey-Level Intelligence

Addressing this challenge requires a fundamental shift in how fraud prevention is conceptualized. Traditional fraud detection is transactional; it examines individual events in isolation and asks whether a payment or login appears suspicious. That model is no longer sufficient.

Fraudsters are patient. They probe. They move across multiple touchpoints, browsing behavior, account creation, password resets, loyalty redemptions, before a fraud event ever reaches a payment terminal. Organisations that are only watching for anomalies at the point of payment are, by definition, watching too late.

The answer is security by design: building continuous intelligence across the entire customer journey, not just at its highest-risk moments. That means mapping intent signals from the very first browse event through to checkout and beyond. It means overlapping coverage across bot detection, account security, scam detection, and payment fraud, because fraudsters don't restrict themselves to a single attack vector, and defenses shouldn't either.

By understanding behavior in context, across sessions, across channels, across time, organizations can distinguish between a legitimate customer using an AI agent to rebook a flight and a fraudster using an automated script to test stolen credentials at scale.

Embracing the New Normal

The emergence of agentic travel is not a risk to be avoided. It is a reality to be navigated. Customers will use AI agents. That traffic will grow. The question for airlines and OTAs is not whether to accommodate it, but how to do so securely and intelligently.

That requires moving beyond binary thinking, human versus bot, legitimate versus fraudulent, toward a richer, more nuanced understanding of intent. It requires platforms that can read signals at the journey level, adapt in real time, and make trust decisions that are accurate enough to protect against sophisticated fraud without creating friction for the growing number of customers and agents acting in good faith.

The gate is open. The agents are already boarding. The organizations that will thrive are those that have built the intelligence to know exactly who, and what, they're letting through.

Darwinium helps organizations understand digital identity and intent across every customer interaction, enabling smarter fraud prevention without compromising experience. Visit Airlines Fraud Prevention