For Kevin Tian, the best solution for artificial intelligence fraud (AI) is also the most obvious: face ia.
In 2022, he founded the Defense Startup against Social Engineering Doppel to achieve precisely that. And as cybercriminals use increasingly advanced models to enhance their attacks, Doppel’s systems have helped companies fight them on a large scale, more quickly and efficiently than before.
The startup has developed AI (software scheduled software to perform specific tasks autonomously) to track Internet, Dark Web and social networks in search of potentially fraudulent activity, detecting all kinds of activities, from imitation websites and false user accounts to malicious ads on Google, Instagram and YouTube. Doppel agents analyze 100 million alerts of this type of phishing threats daily, filtering the real threats of the benign and reporting them to the platforms for elimination. Tian states that they do it with an accuracy of approximately 90% and that they are constantly improved.
“If threat actors can use AI to generate these attacks by only a few cents per dollar, we must ensure that we can also handle that volume on our side,” Tian said Forbes .
On Friday, Doppel announced a new financing round of $ 35 million in a round led by Bessemer Venture Partners. With 55.5 million dollars of risk capital financing, it is now valued at 205 million dollars. This is another milestone for the company that Tian co -founded in 2022 with the director of Technology, Rahul Madduluri, whom he met in Uber while working on the company’s flying cars project, Uber Elevate. Initially, the duo intended to combat fraud related to NFT, helping cryptocurrency companies to track and denounce falsifications. But in 2023, Doppel extended its customer base to other sectors.
In the beginning, Doppel hired contractors in countries such as the Philippines and India to analyze thousands of potential threats and determine which were malicious. But in September 2024, he discovered that the new Openai models, those capable of “reasoning,” could perform the same tasks. He replaced these workers hired by a group of new AI agents and used them to automate 30% of their security operations. Tian states that AI agents have been able to identify more threats than humans. This has been a transformative for the company’s business.
“It is not scalable for a human team to review those millions of alerts every day and now we have an AI agent who can make those decisions,” said Tian.
Also read: This is the Forbes ia 50 list: the main startups that dominate the development of AI
In 2023, notion productivity software faced an avalanche of attacks, including malicious ads aimed at its customers and attempts on social networks to supplant the identity of its executive director, Ivan Zhao. In one case, «[los estafadores] They took the download page of our application, extracted it, cloning it and created a similar domain. In this way, when visiting it, it seemed that you were visiting Notion, ”said Daniel Pyykonen, director of Security of Notion platforms.
So he turned to Doppel in search of help. The company’s artificial intelligence agents quickly solved these campaigns, dismantling thousands. Over time, Notion began to suffer less and less attacks. “We become a very expensive goal,” said Pyykonen.
Doppel’s secret ingredient is a “threat graph”, a map of relationships and interactions between social engineering campaigns (telephone numbers, IP addresses, advertisers accounts). This allows the company to better track malicious hackers through AI as a productivity tool and helps protect future attacks. “We want to offer attackers the same vision so that they can play less to destructive strategy and dismantle all the infrastructure of threat actors,” Tian said.
To date, Doppel has attracted some 180 business clients, including United Airlines, OpenAi and Coinbase, and has tripled its interannual income, according to Tian. It has also incorporated customers in sectors such as oil and gas, finance and insurance, for which it has implemented a second group of AI agents for each sector. “We were surprised by the number of first level clients they had outside the technological sector,” said Elliott Robinson, a partner of Bessemer Venture Partners.
Artificial intelligence has facilitated the process so that cybercounts design new types of social engineering attacks that go far beyond phishing emails with dangerous links. Now they include tactics such as a WhatsApp message from an account that passes through the executive director of a company requesting funds, a false LinkedIn profile that extracts private information and even a realistic lure of the website of a company. In 2024, Google suspended 40 million accounts of malicious advertisers of its platform. Many of them presented images and audios generated by the public figures and executive directors to deceive people and promote scams. Between April 2024 and April 2025, Microsoft stopped fraud attempts that would cost its customers 4000 million dollars and blocked 1.6 million bots record attempts on its Azure platform.
Tian compares the current dilemma that Tom Cruise faced in mission: Impossible – Rekoning : a dishonest system with the mission of distorting reality and controlling the world. “In that movie, AI is bad; it can impeach anyone, create false news and manipulate entire societies simply through digital deceptions,” he said. Consider Doppel as an opportunity to “create a company that solves one of the biggest problems in today’s world: social engineering promoted by AI.”
This article was originally published by Forbes Us.
For more business news, follow Forbes Peru from Google News
Related news :