Posted January 23, 2024

How many fake profiles of you popped up this week? Did your CEO message you asking you to buy gift certificates? Have you ordered a product online only to find the reviews for it were fake? Over the last two years, there has been a rampant acceleration in the number and quality of malicious impersonations online, from celebrities to early stage startup executives. 

What used to be a very manual process for attackers has become highly automated and scalable with the advent of generative AI, turning a fairly rare category of attack into a daily occurrence. In fact, in 2022, imposter scams resulted in more than $2.6 billion in losses in the United States alone. We are seeing daily imposter attacks against a16z, and this is a trend that is happening across all other industries.

Attackers can now replicate your colleagues’ images, voices, and even speech patterns in a very convincing way, with little or no effort. These counterfeiting problems also extend to creating fake merchants, product reviews, and other parts of the online ecosystem. While there are a number of services that historically attempted to handle takedowns, they often leave a lot to be desired. These existing solutions largely rely on a highly manual mix of security operations and legal staff who review information by hand and then follow a long process to remove offending material. 

Enter Doppel. Built by two former Uber software engineers, Kevin Tian and Rahul Madduluri, Doppel is building a next-generation approach for detecting and removing these kinds of threats, focused around automation and AI. Doppel leverages their deep subject matter expertise — Kevin and Rahul helped build key parts of the AI/ML systems that made Uber successful — using AI-native techniques to rapidly detect malicious impersonation, phishing, and disinformation campaigns. We believe Doppel flags suspicious accounts better than their competitors because of the highly customizable tuning within their platform, greatly reducing noise and alerts.

In fact, we’ve seen this situation play out firsthand within a16z. As our firm has grown, so has attention from bad guys, and what was once a rare occurrence has become an hourly problem. Our internal security team spent months looking for a solution to the onslaught of new impersonator accounts created each day. They decided that they needed a tool that took a new approach, embraced automation, and could easily scale. 

The security team’s testing found that Doppel was often 3x faster at detecting and removing threats from popular social networks, and offered coverage on platforms that a lot of the larger incumbents did not. In just a two-week proof-of-concept, Doppel took down about half of what our incumbent solution did in an entire year.

Kevin and Rahul are the kind of innovative and thoughtful founders we love to work with, and this is why we are so excited to partner with Doppel and lead their Series A round.