leveraging the power of graph analytics to analyze complex connections in the IVR through time.
Pindrop® Labs estimates that over 60% of account takeovers start with, or include a call into the IVR
Fraudster calls an institution an average of 5 times before attempting an account takeover attack
How does it work?
IVR activity is analyzed using a patent pending set of technologies analyzing connections between calls and accounts across time, to provide an accurate assessment of which accounts might be under surveillance by a fraudster, if someone is account mining, or spoofing their number to appear as a legitimate customer. These activities and more AI algorithms create a risk score based on the identity claim or account number. That score can be shared across multiple lines of business or enriched with more account data.
When a call with the same number, voice, device, and behavior contacts an account multiple times this is likely what a genuine customer would look like.
If there are unusual calls I.e. ones without the account holders ANI on file. Then, attempting to access the IVR to check a transaction, the spoofed ANI spoof would be caught by Pindrop. The account our spoofing fraudster is accessing is now under attack, therefore its risk rating increases.
Identifying connections between accounts and calls can even be enhanced with other data (*using the same identity claim)
Over time the program starts to identify patterns in larger data sets and predictions get more accurate.
What can you do with account risk?
- Use account risk scores to strengthen security for the most at-risk customers.
- Share account risk with other channels such as mobile, online, in-person, or across LOBs
- Leverage scores to focus precious agent time with increased efficacy
Visualized data snapshot from real IVR with fraud connections appearing in red for demonstrative purposes.
Using artificial intelligence and Trace technology, connections can be analyzed to highlight which accounts are most at-risk, finding attack patterns as far as two months before a takeover is attempted. This provides protection for a customer access point that is often overlooked and underestimated as a target for fraudsters. Through a set of APIs Pindrop can illuminate the most at-risk accounts up to 60 days in advance of account takeover.
Other solutions, whether voice, ANI or behavior based can be limited. They rely on a single factor and are narrowly focused on the call center activity, thus missing the big picture of how fraud is perpetrated across channels. More importantly, other solutions limit themselves to one-to-one relationship between a call and an account and preclude larger and complex multi-level relationships between multiple calls and accounts used by multiple fraudsters operating in tandem. Moreover, other solutions come into the picture while the fraud is taking place or even afterwards; not weeks in advance to help contact centers head the fraudsters off at the pass.
Pindrop’s technology can provide a broader view of the enterprise’s fraud across multiple channels helping to identify fraud that moves between channels. Pindrop is the only provider of graph analytics for the IVR, one of the largest centers for account activity. Connecting this intelligence with your existing data and policies helps make more accurate predictions about risk, and ultimately to catch an increasing number of fraud attempts.