Identity verification with a human touch 

Patrick De Suza, Co-founder and CTO of GRID, doesn’t hold back when voicing his opinions on a variety of topics – from why the effects of fraud on other organisations is often seen as a competitive advantage, to the need for humans in the loop, regardless of how far technology can take us.

We spent time with Patrick to learn more about his thoughts on the identity verification industry today. Here is what he told us. 

Credit bureaus aren’t the be all and end all

Most people believe that credit bureaus are the gold standard, and that all the data that they hold is correct. But the reality is that while they are the aggregators of information, they’re not responsible for ensuring that this information is accurate. In fact, this falls to data furnishers – the entities who report to the credit bureaus. 

Bad actors, meanwhile, have become extremely sophisticated. This especially applies to how they can manipulate data so that it gets past the checks that data furnishes, or lenders, have to report to credit bureaus. 

So many companies assume that when they go to a credit bureau to pull information – that this will be 100% clean and therefore extremely useful. But this isn’t always the case. Instead, it could be manipulated, which will impact your underwriting procedures and your scoring.

Safe as houses

Companies need to think about securing themselves like they would secure their home. 

First, you need a front door – a barrier to entry. You also want to put a lock on that door – maybe even two locks, one at the bottom and one at the top. You might consider putting a camera outside your front door and your windows, as well as motion sensors. Whatever combination you opt for – it’s always going to be a multipronged approach.

No one thing is going to be the ultimate silver bullet in the fight against fraud. To protect your home – to protect your business and your balance sheet, you need to use multiple tools and have multiple steps. That way, that burglar who wants to come in your house – or the bad actor who is trying to get into your business – will be put off by the different security components around your perimeter.

One upmanship

Companies often look at the effects of fraud on other organisations as a competitive advantage. Company A, for example, might find a solution to a particular fraud problem, safe in the knowledge that company B is a long way off solving the same issue. This might sound crazy, but few people equate capitalism with charity. 

But a sensible question might be – what would it take to get companies out of their silos? How can they create joint efforts or consortiums that would foster collaboration and ultimately reduce the impact of fraud on our economy? Might incentives be key here – therefore ensuring that people don’t feel as though they are forgoing their competitive advantage? 

Don’t ignore traditional IDs

There’s a lot of focus on deepfake videos, so much so that standard IDs get missed. 

Today, a fraudster can open a tool and say, ‘Generate an ID for me. I want a driver’s license from X state, saying that my date of birth is X, Y, Z.’ They can provide a photo of themselves, too. 

What is then created is something very real. Tools exist, however, to catch these kinds of ID – telling you that it’s not real, with a list of reasons to back this up. The naked eye wouldn’t pick up on these, but technology can. 

Human in the loop

At present, there’s very much room for manual interaction in the process. We achieve huge efficiencies when we have automation, but that doesn’t mean we should move entirely away from real life input. Instead, it’s about learning when technology needs assistance. 

There’s a misconception that AI and algorithms can do everything. But having a human involved is almost never a bad idea. There are certain applications, for example, where AI will be able to carry out 95% of the job, but that something remains off with the result. And it might be that a human can take a look and make that final determination. 

Technology is great, but it’s not perfect

Manual monitoring and reviewing will help people see blind spots. Innovation remains imperative – no one is disputing that. We should always push the limits. 

But we should also always look to increase the ways in which we are protecting both our companies and our consumers. 

If you’re dealing with millions of customers, you can’t check everything. And doing so probably isn’t a great use of time and resources. But there needs to be monitoring – or process reviewing – in place to ensure that the technology and the innovations that we’ve created are providing the results that we intend them to.