Video Screencast Help
Authentication (User) Blog

Why "Red Flags" would work

Created: 12 Nov 2008 • Updated: 08 Aug 2012
vipblog's picture
0 0 Votes
Login to vote

By Yohai Einav, VeriSign Senior Fraud Researcher The FTC announced last month that is pushing back the deadline for the implementation of the "red-flag" requirements for another six months. Under the "red flags" all financial institutions must develop and implement an "Identity Theft Prevention Program", which includes "reasonable policies and procedures for detecting, preventing and mitigating identity theft". I'm pretty confident that somewhere in the world security chiefs are dancing in relief, and, on the other hand, so are many fraudsters (in their filthy underground caves). FFIEC guidance and beyond So why are fraudsters relieved? Because a well planned and implemented red flag program could actually slow the fraud business. While the 2005 FFIEC regulations (or, "guidance") talked about using better locks to the gates of the castle (which is important, but castles tend to have windows and hidden entrances), the new requirements deal with fighting the enemy within the walls of the castle - inside the compromised accounts. To put it in a less metaphorical way: today, most banks already implement some extra protectional measures at their login page, but only a few measures inside their online banking system itself. And as it seems, better protection of the login - a stronger authentication - does not completely stop fraud, but forces fraudsters to look for the "hidden entrances". (Don't get me wrong - the FFIEC guidance was the cornerstone for all anti online-fraud legislation and the tipping point which propelled anti-online-fraud into the spotlight) Taking care of hidden entrances As it applies to many areas of life, the Pareto principle applies also to the fraud market: 80% of the fraud losses come from 20% of the scam patterns, and a well-thought red flags program will target exactly these 20% of the patterns. Here are a few required red-flags:

  • "Flag an account with a material change in purchasing or spending". This is a strong indicator for financial fraud - someone who suddenly changes his spending behavior - yet today only a handful of financial institutions have applied the mechanisms to detect it;
  • "An account that has been inactive for a reasonably long period of time resumes usage". This is really a common sense red flag, yet only a handful of banks today have the system to detect it.
  • "A consumer report indicates a pattern of activity that is inconsistent with the history and usual pattern of activity of an applicant or member, such as recent and significant increase in the volume of inquiries, or an unusual number of recently established credit relationships". If you learn the behavioral patterns of an account, you could easily be able to find the out-of-pattern activities, and prevent fraud.

Simple? Yes. Effective? Yes, thank you. Would the red-flags policy create a fraud-free environment? No, but it should significantly reduce fraud. Remember the Pareto. And what would be of the fraudsters? It would drive them away from the castle - and back to their filthy underground caves.