Deepfake of CEO with Missing Estonian Accent Exposes $1.5B Veriff to New Cyber Threat

Key facts

• Veriff CEO Kaarel Kotkas was deepfaked in a real-time video call targeting his colleague Andrea Rozenberg.
• The deepfake was detected because the impersonator lacked Kotkas's strong Estonian accent.
• Veriff's 2025 Identity Fraud Report shows a 21% increase in fraud attempts year-over-year.
• Impersonation attacks accounted for 82% of fraud attempts intercepted by Veriff.
• Adversary-in-the-middle attacks surged 46% compared to the previous year.
• Veriff has a $1.5 billion valuation and $192 million in total funding.
• The company processes identity verifications across 230+ countries and 10,000+ document types.

The Deepfake That Almost Worked

Kaarel Kotkas, the 46-year-old founder and CEO of Veriff, narrowly escaped being the face of a sophisticated deepfake scam that targeted one of his senior executives. The attackers used a real-time AI-generated video of Kotkas to request an urgent video call via WhatsApp with Andrea Rozenberg, who leads Veriff's emerging markets. The call appeared authentic — a deepfaked video of Kotkas speaking — but Rozenberg sensed something amiss and verified via Slack, confirming it was a fraud. What gave the impersonation away was not a technical glitch but the absence of Kotkas's strong Estonian accent. "Essentially, the attackers prioritized the visual illusion over the acoustic reality," Kotkas later noted. The incident underscores a growing threat: voice cloning tools are freely available, and real-time video deepfakes are becoming increasingly convincing, making human detection an unreliable last line of defense.

A New Breed of Social Engineering

Kotkas described the attack as unprecedented in its use of genuine relationships and recent shared experiences to lend credibility. "This approach was something completely new," he said. The scammers exploited the trust between Rozenberg and her CEO, crafting a scenario that felt plausible — an urgent request for a call on a relevant topic. The deepfake was not a clumsy phishing email but a targeted, context-aware attack. The incident highlights how fraudsters are leveraging artificial intelligence to bypass traditional security measures. Kotkas warned that organizations can no longer place the burden of detection entirely on individuals. "Hope that AI forgets to clone an Estonian accent is a bad strategy," he remarked, urging companies to shift from checking compliance boxes to proving true human presence and intent online.

Veriff's Expanding Role in Digital Trust

Founded in 2015, Veriff has grown into a global identity verification platform valued at $1.5 billion, with $192 million in funding, including a $100 million Series C led by Tiger Global. The company processes identity verifications across more than 230 countries and supports over 10,000 document types, using AI to match a person to a government-issued ID via a selfie. Its customers include Bolt, Monzo, Wise, Uber, Instacart, Bumble, and Western Union. Kotkas recently announced the acquisition of Vespia, a global Know Your Business (KYB) company, to rebuild KYB from the ground up. The goal is to make business verification 10 times better, 2 to 5 times cheaper, and real-time, eliminating non-value-added fees. The acquisition reunites Kotkas with Julia Ront, Veriff's second employee, who founded Vespia. Veriff also achieved FIDO Alliance Document Authenticity Certification, one of the first IDV providers globally to do so, after testing 19 document types across 12 countries without human intervention.

Fraud Landscape: 21% Surge and New Tactics

Veriff's 2025 Identity Fraud Report, released in November 2024, reveals a 21% increase in fraud attempts overall. Impersonation attacks accounted for more than 82% of the fraud attempts Veriff intercepted, while sophisticated adversary-in-the-middle attacks surged 46% compared to the previous year. Financial services and e-commerce were hit hardest, with authorized fraud in e-commerce reaching over 18 times the global average. highlights regional variability in fraud patterns, underscoring the need for region-specific defenses. Kotkas emphasized the importance of industry-wide collaboration: "Fraudsters share their insights in their communities, unlike businesses in the IDV industry, although we fight the common enemy — fraud online." predicts that artificial intelligence advancements will further shape the fraud landscape in 2025.

From High School Prototype to Global Infrastructure

Kotkas's journey began with a high school prototype that evolved into Veriff, now a cornerstone of digital trust infrastructure. In a conversation with Y Combinator, he described identity as an infrastructure problem and envisioned a future where digital passports might be issued not by governments but by Veriff. The company has verified 400 million people in its first decade and expects to verify the next 400 million in just five months, with every fourth person using Veriff across multiple merchants each quarter. Veriff's growth reflects the broader demand for AI-powered identity verification in regulated industries like fintech, crypto, and gaming. The company competes with Onfido, Jumio, and Persona, but differentiates itself through broad document coverage, high automation rates, and fraud detection models trained on a global dataset of identity fraud attempts. Kotkas remains critical of industry analysts, whom he accuses of promoting "orchestrators" that fail to deliver accurate verification, advising founders to "ignore pay-to-plays and keep on investing every single dollar in R&D."

The Road Ahead: Profitability and the Next Decade

In 2025, Veriff tripled volumes and doubled revenue, surpassing $100 million in revenue while growing profitably. Kotkas admitted the company fell short on two of its three initiatives — delighting customers and building a trust infrastructure for the internet — but praised its global team for creating a winning culture. "When the market says you must choose between conversion or fraud prevention, growth or profitability, work hard or work smart. Our 600-strong global team refuses to compromise," he said. Kotkas's net worth is estimated at $285 million, and he remains based in Estonia, a country known for its digital-first governance. He recently participated in a fireside chat at Slush with Ott Kaukver, his board chairman, discussing the challenges of scaling trust online. As Veriff enters its second decade, the deepfake incident serves as a stark reminder that even the architects of digital trust are not immune to the evolving threats they seek to combat.

The bottom line

• A real-time deepfake of Veriff CEO Kaarel Kotkas was thwarted only because the impersonator lacked his Estonian accent, exposing a new frontier in social engineering.
• Veriff's 2025 Identity Fraud Report documents a 21% rise in fraud attempts, with impersonation attacks making up 82% of intercepted fraud.
• Adversary-in-the-middle attacks increased 46% year-over-year, signaling growing sophistication in fraud tactics.
• Veriff has grown to a $1.5 billion valuation, processing verifications across 230+ countries and serving major clients like Uber and Western Union.
• The company is rebuilding its Know Your Business offering through the acquisition of Vespia, aiming for 10x improvement in speed and cost.
• Kotkas warns that organizations cannot rely on human detection alone and must invest in AI-driven verification to prove true human presence online.