Taking the fight to financial crime through AI

Fighting fraud

Taking the fight to financial crime through AI.


The most successful AI systems are the ones that complement humans. They super-charge our ability to protect people from fraud and financial crime.

Dr Karthik Tadinada

Chief Operating Officer Featurespace



Featurespace’s Adaptive Behavioural Analytics and Automated Deep Behavioural Networks technology uses real-time AI and machine learning software to help banks, insurers and other financial institutions prevent fraud and financial crime.

Fighting fraud with AI

AI technologies, such as machine learning and biometrics, are effective tools to prevent and detect fraud. The technology can monitor and analyse large amounts of data in real time, detecting possible fraudulent activities.

Many financial institutions are now using AI-powered fraud management systems such as that developed by Featurespace, to assist in spotting and preventing scams, fraud and financial crime.

We spoke to Dr Karthik Tadinada, Chief Operating Office, to get his insights into how technology helps reduce risk and fight fraud, particularly cyber fraud and impersonation scams, by helping organisations make informed decisions about which transactions should be authorised.

“The challenge with scams is twofold. Banks need to be able to spot these very risky transactions, and they also need to be able to ask the right questions to get the victim to snap out of the scammer’s spell,” explains Karthik. Featurespace’s AI and machine learning technology enables a financial institution to distinguish good behaviour from bad more easily and make more informed decisions about authorising transactions.

AI’s ‘secret’

Karthik says that the most effective AI systems are those that blend what humans do best with what the computer does best. By using AI to crunch through huge amounts of data at incredible speed, it’s possible to discover and prioritise potentially fraudulent activities quickly. But people then need to review whether or not the activities are truly suspicious and convince customers they’re about to fall victim to a scam.
“I often hear the saying ‘machines are taking our jobs’. But that’s not the case – AI just helps to make us humans more effective.”

Accessible data is key

An AI system is only as effective as the data that feeds it. However, data is often siloed deeply within an organisation, which can make it difficult to know where or how much of it there is. Other potential data gathering hurdles include privacy regulations and data security, which can slow down the process.

For successful AI adoption, organisations need to embrace an open, data-sharing culture: “Being able to make raw data available so that the right people can use it to gain insights, make decisions and provide value is a first-class capability."

Getting the implementation right

Although the implementation process differs slightly from business to business, most of the activities required are broadly similar, including:

1. Systems integration

Establishing the flow of data between an enterprise and the technology platform, working out which decisions are required for each transaction, and thinking through how they’re going to affect the real world.

2. Analytics

Determining what data is needed, how to decide what is and isn’t fraud, and for what business outcomes the system should be optimised.

3. Governance

Project governance and workflow for system users are essential in determining what users need to see, and how they’ll decide what is and isn’t relevant for customers.

This implementation process can take anything from 40 days to 12 months, depending on the number and type of systems and amount of data involved, how much fraud protection is required, how confident the organisation is about AI and how quickly it wants to move.

Irrespective of project size, a successful implementation is dependent on two crucial things: the data and having a single ultimate decision maker who knows what the business is trying to achieve.

Dr Karthik Tadinada

Chief Operating Officer, Featurespace

Protect your business

Technology that helps fight fraud in real time is key to reducing the devastating impact of cyber fraud on businesses. As a global financial institution, Barclays is dedicated to helping protect businesses from the risks of cyber fraud attacks, including helping them stay abreast of the latest threats, and increasing their awareness of what they can do to mitigate the chance of falling victim. To find out more, visit our fraud hub.

“The most successful AI systems are the ones that complement humans. They super-charge our ability to protect people from fraud and financial crime. It’s about making people a lot more effective and guiding their intentions, rather than replacing them.”



Fraud Protection

Fraudsters are working as hard as you are. Our content can help you stay ahead of them.

The power of AI contact form

Contact us

If you're looking to grow your business through AI, our TMT team can help you find the right solution.

See how UK businesses are using AI and machine learning

The power of AI

The power of AI

How companies are using artificial intelligence and machine learning to drive growth.

Game changing AI

Game changing AI

Learn how Peak enables organisations to utilise AI for better decision making, to improve their operations and boost performance.