Deduce is a company that mitigates SuperSynthetic identities, which are AI-generated identities that can pass automated fraud controls, manual reviews, and other identity verification. Pulse 2.0 interviewed Deduce Vice President of Marketing Andy Sheldon to learn more about the company.
Andy Sheldon’s Background
Sheldon is a serial entrepreneur, and this is his third venture-backed startup with an identity graph as its core technology. Sheldon said:
“You might describe me as a one-trick pony. My two previous startups addressed the issue of bot traffic in the media and marketing spaces and were sold to publicly traded companies.”
Formation Of Deduce
How did the idea for the company come together? Sheldon shared:
“To address the issue of fake users, you need a massive identity graph. Scale really matters, and the MAMAA group of companies (Meta, Amazon, Microsoft, Apple, and Alphabet) have a disproportionately large competitive advantage given their installed base of customers. From my experience in the MarTech/AdTech industry, I realized that a massively scaled identity graph could be employed very effectively to address a far more challenging threat of identity fraud.”
Favorite Memory
What has been Sheldon’s favorite memory working for the company so far? Sheldon reflected:
“We are pioneering a solution to an entirely new and rapidly growing threat. Specifically, the use of generative AI by bad actors to create and orchestrate digital identities that appear just like a great new customer to existing fraud prevention solutions. Industry analysts and fraud strategists at some of the largest Fortune 50 companies have recognized that this problem exists and that new thinking is required to combat it, which has been very gratifying.”
“It is really validating of all the hard work we have put in over the past four years and supports our prediction that just like with every technological advancement, such as web 1.0 in 2000 and cloud computing in 2010, the widespread adoption of AI technologies in 2020 would lead to a corresponding jump in fraud.”
Core Products
What are the company’s core products and features? Sheldon explained:
“We have assembled the largest independent identity graph in cyber risk and trust. Today, our graph has seen over 920 million unique email addresses, 331 million online US identities at least once per month, and 185 million unique identities at least three times per week. This gives us a birds-eye view of the online US population.”
“We have developed our own AI-driven pattern recognition technologies, which we call synthetic signatures, to identify identities whose online activities exactly match other identities. Today we are tracking over 845 Signatures comprising over 800,000 SuperSynthetic identities, which are digital identities that are so human-like they will bypass fraud prevention technologies and escalation policies.”
Challenges Faced
What challenges has Sheldon faced in his work sector? Sheldon acknowledged:
“100% of companies we sell to already have solutions in place to prevent identity fraud, so you can imagine it is challenging to get an audience with the fraud team for another fraud prevention tool.”
“We have worked hard to describe this new identity fraud threat and the way that sophisticated criminal enterprises and rogue states are employing AI technologies to create and manipulate identities with the express purpose of defrauding US financial institutions and undermining consumer confidence in the stability of US financial markets and democracy in general. And the problem is getting worse – synthetic fraud accounted for 89% of identity fraud in 2023 and cost US financial institutions over $5 billion in fraud losses.”
Evolution Of Deduce’s Technology
How has the company’s technology evolved since launching? Sheldon noted:
“The scale of our graph has been really impressive. We have grown the graph consistently by 25% annually. The addition of confirmed fraud data from our customers has significantly improved our generalized scoring model. Our graph typically recognizes 90% of our customers’ new users during the new account opening workflow.”
“We typically see a 5-7% improvement in fraud capture with multiple incumbent vendors in place and dramatically reduce the false positive rate (i.e., misidentifying a fraudster or good customer over legacy solutions). This is why we have adopted a channel-based go-to-market strategy to partner with incumbent vendors as an incremental, complementary solution to their own products.”
Significant Milestones
What have been some of Deduce’s most significant milestones? Sheldon cited:
“Obviously, raising capital along our journey has been important. We raised $9 million in 2023 in what was a brutal financial market that saw many early-stage companies shutter. This is because we have a very supportive group of investors and board who see the value that Deduce is bringing to financial markets and the differentiation that is being realized now by our channel partners.”
Total Addressable Market
What total addressable market (TAM) size is the company pursuing? Sheldon assessed that the identity affirmation market is $85 billion.
Differentiation From The Competition
What differentiates the company from its competition? Sheldon affirmed:
“There is a systemic problem in the way that identity fraud is being addressed today. This has worked historically because of the breadth and number of databases and vendors with identity data. This new AI-generated fraud threat uses data that is already in databases, such as credit history and email reputation scores.”
“Consequently, when the fraud platform queries their multiple data sources for insights on a specific identity, the returning information describes an ideal customer. Deduce compares an individual identity against all other identities in our graph to detect patterns of activity that suggest sophisticated synthetic identities. Therefore, scale and our proprietary use of AI-driven pattern recognition technologies are what differentiates us.”
Future Company Goals
What are some of the company’s future company goals? Sheldon pointed out:
“Revenue is the lifeblood of any business, so we are laser-focused on driving business through our channel partners. We are describing a new category of identity fraud so it is essential we continue to educate the market about this new threat and how to detect sophisticated synthetic identities at new account openings.”
“The even bigger threat are the identities that have already been onboarded and are now funding their accounts and checking account balances with the express purpose of appearing to the bank’s risk algorithm as a customer in good standing and therefore worthy of credit. This results in a rise in first-party synthetic fraud. Deduce will introduce a product that will provide data science teams with a data source to help identify fake customers before they become a fraud loss.”
Additional Thoughts
Any other topics to discuss? Sheldon concluded:
“While we are focused specifically on financial services and fintech markets and helping our customers identify fake customers, it is clear that fake humans are ruining society for real humans. While we applaud the Biden administration’s regulation on AI, it does not go far enough to address the authenticity of online identities to prevent bad actors from using their fake identities to impact everyday life for real humans, starting with our basic demographic principles and elections.”
