Deepfake competition from FinVolution seeks innovative detection models

Deepfake detection is in the spotlight at the FinVolution Group’s 2025 FinVolution Global Data Science Competition.

A release says the tenth edition of the contest challenges AI researchers, engineers and data scientists worldwide to develop innovative models that can “identify and counter manipulated visual content.”

In other words: detect deepfakes.

Tiezheng Li, CEO of FinVolution Group, says that “by focusing on deepfake detection this year, we aim to advance AI safety while encouraging open innovation that benefits society at large.”

The competition is integrated with the Conference on Information and Knowledge Management (CIKM) and the International Joint Conference on Artificial Intelligence (IJCAI).

The prize pool is 308,000 renminbi (approximately US$42,900), and winners will receive an invitation to showcase their work at CIKM 2025 in November.

The preliminary round runs until July 23, followed by a semifinal round in August and a final challenge in early September.

Deepfakes ‘an insidious threat’ to financial services: Veriff

The scale and breadth of the deepfake threat continues to grow. Ira Bondar, senior fraud group manager at Veriff, says that in the first quarter of 2025 alone, “the number of reported deepfake fraud incidents nearly matched the total from 2024. Deepfakes are an insidious threat, and we are now officially in an age when it’s becoming nearly impossible for the naked eye to distinguish what’s real from what’s not, creating great peril for both financial services firms and consumers.”

Bondar’s comments come as Veriff launches its latest annual report, “The Future of Finance.” Key findings from the report show 60 percent of firms reporting an increase in the use of AI in attacks, and 89 percent of senior fraud decision-makers in financial services predicting a continued increase in fraud through 2025.

Moreover, “82 percent of consumers will not sign up on a financial services platform if they are not confident in the organization’s fraud defenses; and 38 percent of consumers picked biometrics as their number one choice to protect their interactions, with facial recognition being the most popular biometric option.”

Bondar says the report “collates a wide variety of research to present a clear picture of how financial services organizations – from traditional banks, to agile neobanks and fast-growing cryptocurrency, fintech and payments platforms – should consider approaching the daunting challenges of the coming years.”

For Bondar and for Veriff, the door to customer satisfaction, security and compliance for financial services firms can be opened – and “identity verification, biometrics and AI hold the keys.”

Article Topics

biometrics  |  deepfake detection  |  deepfakes  |  FinVolution Global Data Science Competition  |  fraud prevention  |  research and development  |  Veriff