AI News caught up with Siddhartha Choudhury, a Senior Product Manager at Booking.com, to get the inside scoop on how the technology is keeping your bookings – and data – safe from security threats like online fraud.
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 assistanceThe 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.
Headshot for Siddhartha Choudhury, a Senior Product Manager at Booking.com, for an article on how the technology is keeping your bookings – and data – safe from security threats like online fraud.“We use AI for a broad range of safety and fraud risk mitigation use cases,” explained Choudhury. “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.
In short, they combine the best off-the-shelf software with their own custom-built AI to create a powerful security cocktail, protecting both travellers and property owners on the platform.
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 threatsIt’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.