Fujitsu looks towards social media in the fight against fraud

Identity will be the cornerstone of the next generation of the internet and as more transactions move online, new and sophisticated reputation systems will consolidate our currently balkanised digital selves. Today we see financial institutions are increasingly looking to things like social media to scrutinise for potentially fraudulent behaviour.

An example where we are starting to see this happen is in insurance claims where investigators can look at social media profiles of claimants to try and detect patterns of fraudulent behaviour. Anthony Duffy, director retail banking at Fujitsu, who has 30 years' experience working in fintech, pointed to this trend. If someone prangs their car and the claim turns out to be fraudulent, and then someone they are linked to on Facebook also crashes their car and makes a claim, this would be up for scrutiny, he said.

"It doesn't mean to say the person who is linked on Facebook is making a fraudulent claim. But because they are linked this might mean a bit more time was spent investigating that claim to see if we are both running some sort of scam. That sort of intelligent use of data is becoming more common."

This is an interesting, evolving area within financial technology. Walk around the recent Finnovate 2017 event in London and you could chat to a number of teams from on-boarding solutions all expounding the possibilities to be had within social media. But it's an uncharted terrain that throws up host of challenges as well as new privacy issues.

In his book, Identity is the New Money, David Birch cites a lawsuit filed in the US where a cheerleading coach demanded members of her squad hand over their Facebook login information after some indiscrete postings were shared among students. Faced with this privacy violation the students deleted their profiles; users not only understood when their privacy was being compromised, efforts to prevent them manipulating their virtual identities proved fruitless.

Fujitsu works with a large central government department that supports welfare payments, and social media profiles also came into play during the investigation of benefits fraud. Duffy recounted how fraud involving twins and child support was picked up in the Liverpool area.

"There seemed to be an outbreak of twins around Merseyside. Someone had spotted that if you telephoned the local government office saying there had been some mistake with an application for child support; they were telling the benefits office that in fact they had two children who were born on the same day, but were only receiving money for one."

In this case the benefits office was collecting additional details on the telephone and adding it to the system and money was then paid. "The fraudulent claimants had worked out there was no additional security checks being done. After that word must have got around the school playground and the fraud was duplicated."

There have been great advances with the way banks can quickly check the legitimacy of driving licences, for instance, by reading security measures on the physical documents and also linking to the DVLA database. However, they would struggle when presented with foreign IDs, said Duffy.

"Think about the number of people from overseas who came into the UK in the last three years. If someone walked up to me with a Polish photocard I wouldn't know whether that was a bus pass or a national ID card. Banks struggle with that so this sort of system where you can actually link in and prove identity is quite an interesting area."

Fujitsu also offers some nice biometrics solutions, including palm vein image verification. "There are 36 points we take information about where blood flows into the hand and capture this information and embed it on a high-end chip on a plastic card. When the card goes into an ATM, the image of the hand that is presented is matched against the image on the card."