I think of this as a multi-layered, multipronged approach. You want to have an “at the front door” protection system, like your security system at the front door, to prevent bad actors from even entering in the first place. That's where we use something we call “advertiser verification”. We apply an onboarding trust assessment. What we're trying to do there is assess the veracity of the actor on our systems and then triage them into either a standard or enhanced verification process. For something like financial services, that is considered a higher-risk sector, so they would go through an enhanced verification process.
In addition to that, we would have our systems working at the back end, constantly analyzing their behavioural signals so that we can understand what the risk profile of this actor is.
There are a whole host of systems that are deployed, including in the onboarding process, to ensure that we have the right people—