Recently, DDC signed a new vendor partner: Doppel. We were excited to add Doppel to our catalog because it addresses a significant security concern that is top of mind for lots of campaigns this election cycle: Impersonations of campaigns and the use of inauthentic content about candidates.
We had a chance to pose these questions to Doppel’s founder, Kevin Tian, about the risks campaigns face and how Doppel addresses them.
DDC: Tell us about how Doppel was started and what led you to launch the company.
Rahul and I met as engineers at Uber working on AI, ML, and distributed systems powering the core ridesharing platform. We wanted to start a company together and made it happen two years ago.
Doppel started off as an AI tool for detecting scams in crypto, the most fast-paced adversarial threat environment. From there, we realized the technology we built to scale for volume and velocity was very valuable for serving multinational enterprises, so we expanded to cover all digital attack surfaces, including social media, domains, and telco.
DDC: There is a lot of noise about deep fakes, inauthentic content, and impersonators. What is your view of the threat landscape and is the situation getting worse? If so, why?
While there certainly is a lot of hype, the threat actors and the tactics they deploy pose very real threats to businesses, organizations, and campaigns. AI, deep fakes, and disinformation accelerate existing impersonation threats by making it cheaper and more personalized. Just take a look at the news, every day we see new examples of hyper-personalized and targeted social engineering campaigns that manipulate public opinion, create physical threats, and drive rampant financial loss to victims.
Year over year, the number of deepfakes detected has grown exponentially—we've seen estimates of 90,000 deepfake videos shared online in 2023; when including synthetic audio, that number jumps to ~500,000, according to sources like DeepMedia.
DDC: In our discussions leading up to you offering Doppel to DDC-eligible entities, we sensed a lot of passion for helping campaigns. Why was it so important for you and the company to provide this protection for free?
At Doppel, we’re on a mission to make the internet a safer place. With pivotal elections this year, we have the tools to help the electorate and want to do our part to not only raise awareness of the threat but also play a role in protecting society from serious risk.
DDC: Campaigns have little time for scouring the internet and keeping on top of efforts by bad actors' activities about them on social media. Can you give a little insight into how your product works? What does it do and how does it do it?
Doppel leverages AI technology and security experts to automate the collection, validation, and takedown of social media threats. We crawl the web for potential threats, capture evidence of potential threats, and use AI to categorize threats. From there, our security experts validate and take action on the threats to protect clients from these bad actors. It’s turnkey and only requires a brief onboarding call to get started.
DDC: Campaigns don’t have a lot of time to track and request takedowns of false content. How does your takedown process work? What are the limitations around takedowns (or maybe something around success or how social media sites evaluate? Just want to create some realistic expectations)
Our takedown process involves working directly with the platforms to take down content based on their policies. When we validate that a threat violates the platform policy, we have an extremely high success rate, over 95%. However, if the threat does not violate a policy, we would not be able to take it down. For example, parody accounts are often protected by free speech principles on the platform.
DDC: Just because you request a takedown, doesn’t mean the social media companies are obligated to do it. How does that work and what should the expectations of campaigns be vis a vis takedowns?
Our efficacy rate is high when we can collect evidence that there is a clear violation of their terms. Typically, platforms have policies against direct impersonation, and these pose the largest threats from a cyber risk posture.
DDC: Do you have any success stories you can share?
Absolutely. Our commercial customers include major Fortune 500 companies like Meta and Coinbase, financial services firms like Ark Invest, to Hollywood talent agencies. We help them prevent their executives from getting impersonated and reduce social engineering risks to their business.
DDC: Any final thoughts?
We’re excited to work with DDC and offer our technology to protect election integrity. The threat landscape is rapidly evolving with AI, and we believe cutting-edge problems require cutting-edge solutions. Looking forward to partnering with campaigns and fighting the good fight together!
If your campaign is eligible to receive DDC products, contact us at info@defendcampaigns.org