Matthew RosenquistFounder & CISO Advisor, Cybersecurity Insights · CISO, Cybersecurity Strategist, & Industry Advisor

Identity Fraud Report Highlights Rise in AI Attacks

16 Feb 2026 · 3 min

Watch on YouTube

2:37

In this video, Matthew Rosenquist summarizes SumSub’s 2025‑2026 Identity Fraud Report, which documents a rapid escalation in the sophistication and volume of identity‑theft attacks driven by artificial intelligence. He explains that while defenders have made progress—through broader adoption of multifactor authentication, better fraud‑education programs, and more effective phishing filters—cyber‑criminals are increasingly leveraging AI to produce high‑quality, scalable fraud. Key findings include a 180 % increase in fraud‑attempt quality, the share of advanced attacks rising from roughly 10 % in 2024 to 28 % in 2025, and 40 % of companies and 52 % of end‑users reporting at least one fraud incident. The report details AI‑enabled tactics such as synthetic identities, layered social‑engineering, device/telemetry tampering, and the creation of realistic fake documents, images, videos, and deep‑fake impersonations. It also breaks down trends by country, region, industry, and attack type. Viewers are encouraged to download the full report via a link in the description.

Chapters

Digital fraud is on the rise and we're entering into a new phase where attackers are leveraging AI to elevate their sophistication and increase their effectiveness. My partners at SumSub have just released their 2025-2026 Identity Fraud Report that highlights this race between attackers and defenders. The report showcases the struggle On the one side, fraud education is getting better, multifactor authentication controls are being widely adopted, and improved phishing filters are doing a great job at undermining low-effort attacks such as shoddy document forging, simple impersonations, and generic phishing campaigns. But the cybercriminals are motivated persistent and adapting. They are fully embracing artificial intelligence tools to raise their game and create sophisticated, industrialized, and scalable fraud that is harder to detect and much greater in quantity. The numbers are impressive. Fraud attempt quality is up 180%, showcasing attacks that combine multiple coordinated techniques such as synthetic identities, layered social engineering engagements and device or telemetry tampering. In 2024, only about 10% of fraud attempts were advanced, but in 2025, it has surged to 28%. Overall, 40% of companies and 52% of end users reported experiencing fraud at least once. AI's pursuit for realism is a powerful capability for the attackers. Fake documents, images, videos, impersonations, and forgeries, just to name a few. AI tools and agents are providing attackers a new level of believability for their fake identity documents, social posts, autonomous conversations, video interactions, impersonations, and Forgeries. SumSub's report breaks down the fraud trends by country, region, targeted industries, and attack type activity. Download the full report and see how fraud sophistication is on the rise. The link is in the description.