n early 2024, an employee of a global financial organization unintentionally wired $25.6 million to fraudsters.¹ The employee, under the impression they were talking to the company’s CFO and other senior leaders of the company on a video call, was maliciously deceived by deepfake technology. The fraudsters used deepfakes of the CFO and senior leaders to simulate their likeness and gain the employee’s trust before collecting their payout.

Deepfakes are forged or digitally-altered media created by generative artificial intelligence (AI) designed to impersonate people and events. Despite their widespread use for entertainment on social media, deepfakes have emerged as a growing source of loss in cyber insurance² and pose significant risks to insurance companies. Between 2022 and 2023, Allianz reported a 300% increase in doctored claims photos.³ In a recent study from Verisk, nearly all (98%) of insurers agreed that AI-powered editing tools are fueling an increase in digital insurance fraud.⁴ Insurance fraud has not only become more frequent, but also harder to detect due to the increased availability and sophistication of AI tools. About 50% of Gen Z and millennial consumers reported being “at least somewhat likely” to make a small edit of a claim photo or document, while only 32% of insurers say they are “very confident” in detecting deepfakes. ⁴

Given their far-reaching implications, deepfake incidents could be covered under cyber liability, directors and officers liability, errors and omissions liability, commercial general liability, and umbrella insurance. Potential consequences of deepfakes include defamation, misinformation, identity theft, blackmail, and financial fraud.⁵ As it stands today, there is no one set rule for where deepfake incidents belong. Slowly but surely, coverage options and endorsements have started creeping into the insurance marketplace to fill this gap. Coalition, a tech-driven cyber insurer, offers coverage specifically for deepfake-related incidents.⁶ Chubb, a leading traditional cyber insurer, has also added targeted coverage for deepfakes through their social engineering fraud endorsement.⁷

Still, there is an immense opportunity for actuaries to design insurance solutions for the $15.3 billion cyber insurance industry — and fast. According to the FBI, cyber insurance losses and fraud scams increased by 33% from 2023 to 2024.⁸ Now more than ever, actuaries can play a key role in staying ahead of evolving attack vectors through innovative product design and quantifying exposure and development potential.

Actuaries not directly involved in cyber insurance must also stay vigilant. The Coalition Against Insurance Fraud (CAIF) estimates U.S. insurers pay over $300 billion each year in fraudulent claims, with one in ten property-casualty losses found fraudulent.⁹ Many insurers use third-party and internal AI-based detection tools, while some require additional claim documentation metadata analysis (timestamps, location, etc.) before a claim payout.¹⁰ Yet, advances in AI tool capabilities, combined with creative consumer tactics, seem to continuously outpace insurers’ fraud-detection strategies. Devising better ways to detect fraudulent media remains a priority, and actuaries can use their broad purview to advocate for strong data governance to enable the full potential of modern anti-fraud tools.

On the bright side, on March 10, 2026, Zoom launched a deepfake detection feature for live video meetings.¹¹ Hopefully this prevents any actuary from becoming the subject of the next deepfake-induced corporate fraud incident.