Introduction to Online Security
When you book a holiday online, you’re placing a lot of trust in a website. You trust it with your money, your personal details, and your travel plans. For a giant like Booking.com, keeping that trust for millions of people every single day is a huge job. So, how do they do it? Increasingly, the answer is AI.
Countering Modern Online Fraud Requires AI Assistance
The sheer amount of data Booking.com handles is hard to wrap your head around. This isn’t just about stopping someone from using a stolen credit card; the platform has to spot everything from fake hotel reviews and marketing scams to phishing attacks and account takeovers.
“We use AI for a broad range of safety and fraud risk mitigation use cases,” explained Siddhartha Choudhury, a Senior Product Manager at Booking.com. “Here, we deal with petabytes of data which includes events generated from applications, infrastructure, messages, emails…”
To handle this, they don’t rely on a single magic tool. “We also leverage multiple vendor specific built-in ML solutions and in-house solutions together to identify and mitigate fraudulent attacks,” Choudhury adds.
The Million-Dollar Question: Better or Cheaper
Naturally, running a security system at this scale isn’t easy. One of the biggest headaches is simply getting all the different tools, both internal and external, to play nicely together. But Choudhury pointed to an even tougher, more constant balancing act: performance versus cost.
Cyberattacks get smarter every day, which means your defences constantly need to get better. But better tech costs more money.
“Due to evolving cyber threats, attacks are more sophisticated and the scale of data is increasing constantly,” says Choudhury. “So the decision is: should we make things more cost-efficient, or should we need to make it even better performance wise?”
How AI Helps to Get Ahead of Online Fraud Threats
It’s often said that the best defence is a good offence. Instead of just reacting to problems after they happen, Booking.com is using AI to spot trouble before it even starts. A big part of this involved moving their systems to the cloud, which allows for smarter and faster tools.
Choudhury explained that their human security experts now have a team of digital helpers. “Multiple AI assistants are working in parallel for security analysts to improve their efficiency and reduce operational toil,” he explains.
By giving your best human detectives a digital partner that can sort through mountains of clues in seconds, the experts can focus their skills on the most critical threats, while strong monitoring systems make sure the AI itself is running smoothly and accurately.
Making Sure AI Plays Fair
When you give AI the power to make important security decisions, you have to be incredibly careful that it doesn’t become unfair. Choudhury was clear that ethics are at the core of their strategy, which is built on a few key ideas:
- Fairness: The company actively checks their AI for biases to make sure it’s not unfairly flagging certain people or groups for potential online fraud.
- Human oversight: This is a big one. For any major decision, “there is enough human involvement to identify false positives.” A person, not just a program, often gets the final say.
- Explainability: The AI can’t be a mysterious ‘black box.’ The team needs to understand why it made a certain decision to ensure accountability.
- Privacy: And, of course, everything is built on a solid foundation of “strong data protection, privacy and compliance,” always with the user’s consent in mind.
What’s on the Horizon
Choudhury believes the next big leap isn’t about finding brand new things for AI to do, but about making all the existing AI tools work together efficiently.
“I expect more and more solutions will be designed but also expect them to be orchestrated in a way that will make departments much more efficient,” Choudhury predicts.
The goal is to build a system where all the security parts communicate and collaborate intelligently. For Booking.com, the mission is clear: “Driving innovation alongside guaranteeing reliability and cost efficiency will be the key focus.”
Conclusion
For the rest of us, AI is giving us a little more peace of mind we won’t become victims of online fraud after clicking ‘book’ on that holiday. Siddhartha Choudhury and the Booking.com team will be sharing more insights at this year’s AI & Big Data Expo Europe in Amsterdam on 24-25 September 2025.
FAQs
Q: How does Booking.com use AI to prevent online fraud?
A: Booking.com uses AI for a broad range of safety and fraud risk mitigation use cases, including identifying and mitigating fraudulent attacks.
Q: What are the key challenges in implementing AI-powered security systems?
A: The key challenges include balancing performance and cost, as well as ensuring that the AI system is fair, transparent, and accountable.
Q: How does Booking.com ensure that its AI system is fair and unbiased?
A: Booking.com actively checks its AI for biases and ensures that there is enough human involvement to identify false positives.
Q: What is the future of AI-powered security systems in the online booking industry?
A: The future involves making all the existing AI tools work together efficiently, driving innovation alongside guaranteeing reliability and cost efficiency.
