
Baran Ozkan spent 15 months looking for software that did not exist. As head of financial-crime product at a European bank, he wanted a single tool to catch money laundering and fraud. Vendors overpromised. His own employer tried to build one in-house and failed.
So he built it himself. His startup, Flagright, has now raised a $12.5mn Series A led by Infinity Ventures, a San Francisco fund whose partners ran deals at PayPal. Sella Direct Ventures joined as a new backer, alongside existing investors Frontline Ventures and Y Combinator.
The ‘operating system’ pitch
Flagright calls itself the AI operating system for financial-crime compliance. In plain terms, it tries to replace two things at once: the legacy systems banks have run for years, and the patchwork of separate tools bolted on top.
It folds transaction monitoring, watchlist screening, risk scoring, case management and investigations into one platform. When a payment moves through a customer’s system, Flagright checks it in real time for fraud, money laundering and sanctions risk. It then hands compliance teams a result they can explain to a regulator.
That last point matters. Banks will not deploy a black box. So Flagright leans hard on “explainable” AI, with auditable agents and humans kept in the loop. Clients replacing older tools report up to 93 per cent fewer false positives, the company says.
AI versus AI
The timing is the pitch. Banks are squeezed from two sides: more transactions to watch, and criminals using AI to industrialise fraud. Interpol puts global fraud losses at $442bn, and says AI makes the crime far more profitable.
Defenders are answering with their own AI, and investors are funding it hard, from Behavox’s $175mn raise to startups automating regulated back-office work. The financial-crime compliance market was worth about $26.5bn in 2025 and could nearly triple by 2034, by one estimate.
A crowded field, a narrow window
Flagright is small: about 40 staff, offices in London, San Francisco and Singapore, and more than 100 banks and fintechs across 30-plus countries. It is up against deep-pocketed incumbents such as NICE Actimize, Feedzai and ComplyAdvantage.
Ozkan’s case is timing. Older players are bolting generative AI onto old systems, he argues, while Flagright built on it from the start. “We are the oldest company for generative AI technology in financial crime,” he claims, an odd boast for a three-year-old, but one that captures how slowly the industry moved.
The new money goes to sharpening the AI and pushing into the US. The real question is whether a 40-person startup can set the standard before the giants catch up.
View original source — The Next Web ↗


