Catching AI fraud with behavioral profiling

This article is a part of Startup Spotlight, a series that features young, up-and-coming startups.

Image credit: Timmy Loen

Scammers used founder Julia Jakimenko’s cloned face and voice to target elderly victims in a fraud scheme. Authorities were unable to pursue the case because investigators lacked the forensic evidence needed to identify those responsible.

That gap led Jakimenko to launch Cyberette, a platform designed to help organizations investigate AI-generated fraud through media forensics and digital evidence analysis.

😟 Problem

Deepfake fraud cost global businesses US$25 billion in 2024. According to the FBI, 90% of digital evidence in cybercrime cases suffers from integrity issues. Current defense tools only identify fake media without providing the forensic data required for prosecution.

💡 Solution

Cyberette operates a forensic infrastructure platform that detects AI-generated deception in real time. The software provides evidence and contextual analysis needed to investigate deepfake-related fraud.

The software utilizes several core technical capabilities:

• Real-time detection of synthetic voices, faces, and manipulated behaviors.
• Multilayered behavioral analysis to identify signs of distress or coercion.
• Automated packaging of forensic evidence to support legal prosecution.

Image credit: Cyberette

📊 Market size

The startup operates within a US$50 billion serviceable addressable market spanning defense, intelligence, and enterprise security. Cyberette targets a US$400 million reachable market within three years.

Demand for these tools is increasing as organizations face rising levels of AI-generated fraud, identity scams, and regulatory scrutiny around digital evidence.

🤝 Team

• Julia Jakimenko. Founder and CEO. She brings experience from previous roles at Coalition for Content Provenance and Authenticity and Rabobank.
• Bryan Saldivar. Data scientist. He works on managing large-scale datasets and improving AI model performance and deployment.
• Jeremy King. Broad advisor. He brings more than 25 years of experience in fundraising, investing, and global investor relations.

🚀 Traction

• Maintains detection accuracy above 93% across image, video, and audio formats.
• Secured four global partners across South Korea and Singapore.
• Acquired customers across intelligence and corporate security in seven countries, converting 85% of pilot programs.
• Accepted into the Nvidia Inception program and supported by Microsoft and AWS startup initiatives.
• Selected as a top 50 finalist in SuperAI’s Genesis Startup Competition.

🏆 Competition

💰 Financials

🚩 Risks