Sumsub Debuts AI Copilot to Transform Compliance Workflows and Fraud Investigations

WHY THIS MATTERS:
Compliance and fraud teams are under mounting pressure from two converging forces: rapidly escalating AI-powered fraud and tightening global regulation. Sumsub’s own data shows AI-driven, multi-step fraud schemes rising 180% year-on-year in 2025, while frameworks such as the EU AI Act and Singapore’s Protection from Scams Act are increasing expectations around explainability, documentation and human oversight.

The result is a structural imbalance: fraud is becoming faster and more automated, while compliance obligations are becoming more detailed and evidence-driven. Traditional tooling—dashboards, static rules engines and manual case reviews—is struggling to keep pace with both volumes and complexity.

Summy’s evolution from an AI assistant within case management to a platform-wide AI Copilot reflects a broader industry shift: embedding AI directly inside regulated workflows rather than layering it on top. By grounding outputs in platform-native data and keeping final decision authority with humans, Sumsub is positioning Summy as an augmentation layer rather than a black-box decision engine—an important distinction in regulated markets.

Sumsub, a global verification and fraud prevention leader, today announced the launch of its new Summy AI Copilot (“Summy”), the first platform-native AI agent for compliance and fraud teams. Fully integrated into the Sumsub platform, Summy transforms complex, real-time case data into clear, actionable insights within existing workflows, enabling investigators and compliance officers to make faster, smarter decisions without leaving the platform. 

Summy was first introduced as an AI Assistant within Sumsub’s Case Management solution last year, where it equipped financial crime teams to manage growing alert volumes by prioritizing key risk signals and shortening investigation cycles. The new release evolves Summy from a case-focused assistant into an AI Copilot spanning the entire Sumsub platform, now supporting a broader set of compliance and fraud workflows.  

This launch comes amid a surge in sophisticated fraud attacks. According to Sumsub’s Identity Fraud Report 2025-2026, multi-step, resource‑intensive schemes powered by AI increased 180% year‑on‑year globally in 2025. At the same time, global regulators are tightening anti‑scam and AI‑related rules—from the EU Artificial Intelligence Act to Singapore’s Protection from Scams Act—raising expectations for transparency and documentation, while pressing compliance teams to make well-reasoned decisions in a more timely manner.

Embedded across the Sumsub platform via an AI chat interface, Summy is a large language model‑based AI Copilot that works entirely through plain‑language queries. Designed for compliance officers, risk managers, and fraud investigators, it delivers precise, audit‑ready outputs grounded in Sumsub data, ensuring decisions are traceable and well-documented within existing workflows. Operating within thresholds and controls set by compliance teams, Summy keeps AI-driven actions aligned with established policies, eliminating black-box automation and ensuring decisions remain under human control. In practice, it significantly reduces manual analysis and accelerates case handling, helping risk and compliance teams achieve productivity gains of up to three times, on average.

Key capabilities of Summy include: 

• Product knowledge: Accessible 24/7 through the AI chat interface, Summy instantly retrieves information on Sumsub’s features, setup flows, and troubleshooting, eliminating delays in obtaining technical support. 
• Visual analytics: Generates charts and graphs from platform data to highlight trends and performance insights in seconds for compliance and fraud teams. 
• Compliance advice: Provides structured, market-specific guidance on regulatory requirements and workflow design, aligned with the latest rules and expectations. 
• Case summaries: Summarises complex cases and alerts into concise overviews that highlight key risk signals and suggested next steps, supporting quicker and more consistent decisions.  

“Summy has become an integral part of our daily operations, providing our team with rapid clarity on compliance workflows and product usage without relying on lengthy documentation or manual support requests,” said Svetlana Shchennikova, KYC Product Ops Lead, Mercuryo. “The AI Copilot frees up capacity for deeper analysis of more complex, high-impact decisions. While human expertise remains central to our process, Summy has proven to be a reliable first point of reference for compliance and product inquiries.” 

“Unlike autonomous AI systems that make opaque decisions, Summy is designed as a supporting AI agent for compliance teams, not a substitute for them. All decisions remain under human control, without sacrificing reliability, traceability, or accountability,” said Andrew Novoselsky, Chief Product Officer at Sumsub. “With Summy and our recently launched AI Agent Verification, we are investing in an AI ecosystem where agents are explainable, tied to real accountability, and built to help businesses stay ahead of increasingly sophisticated fraud.” 

FF NEWS TAKE:
This launch underscores the next phase of AI adoption in financial crime prevention: operational copilots, not autonomous decision-makers.

The market has moved beyond experimentation with generative AI chatbots. Institutions now need production-grade AI that is explainable, auditable and embedded within compliance architecture. By focusing on traceability, policy-aligned controls and audit-ready outputs, Sumsub is clearly targeting regulated enterprises wary of regulatory backlash from opaque AI systems.

If tools like Summy consistently deliver measurable productivity gains—Sumsub cites up to 3x improvements in case handling—AI copilots could become standard infrastructure for compliance teams, much like CRM systems became indispensable for sales.

The real competitive battleground will be trust: which providers can offer AI acceleration without compromising governance. Sumsub is making a clear bet that explainable, human-supervised AI—not autonomous enforcement—is the model regulators and enterprises will accept.