Sumsub and ODDA Urge Dating Apps to Combat AI-Driven Fraud

A new white paper from verification platform Sumsub and the Online Dating and Discovery Association (ODDA) addresses the growing issue of romance fraud in the online dating sector, particularly as AI technologies make deception more sophisticated. The report analyzes how AI-generated content, deepfakes, and synthetic identities are complicating trust and safety efforts across dating platforms. It provides practical recommendations for improving fraud prevention without undermining the user experience that makes these platforms attractive.

The Growing Threat of AI-Powered Romance Fraud

The report, "Are You Real? How Dating Apps Must Re-Architect for the AI Fraud Era," analyzes how AI-generated content, deepfakes, synthetic identities, and coordinated scam networks are complicating trust and safety efforts. A major focus is on the difficulty platforms face in verifying genuine users while maintaining a positive user experience. This challenge has only intensified thanks to the widespread accessibility of AI image, voice, and video generators.

The paper draws on Sumsub's fraud prevention data and ODDA's industry expertise to examine the scale of romance fraud. It provides detailed analysis of how AI tools are evolving tactics used by bad actors. The research includes survivor perspectives and industry data that illustrate the human and financial costs of these sophisticated scams.

Key sections of the report cover the current regulatory environment, including implications of the Online Safety Act. The analysis extends to broader fraud strategies in the UK and elsewhere. Understanding these regulatory frameworks is essential for platforms seeking to implement effective safety measures.

Limitations of Current Safety Approaches

The report discusses significant limitations in existing trust and safety approaches currently deployed by dating platforms. One-time identity checks and reactive moderation are proving insufficient against evolving threats. These traditional methods often catch fraud only after damage has occurred, leaving users vulnerable to sophisticated scams.

Current trust and safety measures like one-time identity checks and reactive moderation are proving inadequate against evolving AI-powered scam tactics.

Bad actors are leveraging AI to create convincing fake profiles that can bypass initial verification steps. Synthetic identities generated through AI tools are becoming increasingly difficult to distinguish from genuine users. The speed at which scammers can now create and deploy fraudulent accounts has outpaced traditional detection methods.

The DATE Framework: A Structured Response

To address these challenges, the paper proposes a DATE framework as a structured response for platforms. The framework focuses on four key pillars: risk-based deterrence, behavioral analysis, high-risk triage, and user empowerment. This comprehensive approach aims to create multiple layers of protection that work together to identify and prevent fraud.

Risk-based deterrence involves implementing verification measures proportional to the potential threat level of different user behaviors. Behavioral analysis examines patterns over time rather than relying solely on initial verification. High-risk triage allows platforms to focus resources on the most suspicious activities and accounts.

The DATE framework focusing on deterrence, behavioral analysis, triage, and empowerment provides platforms with a structured approach to combat AI-enabled fraud.

User empowerment represents a critical component of the framework, giving members tools to protect themselves. This includes education about common scam tactics and clear reporting mechanisms. By involving users in the safety ecosystem, platforms can create a more resilient defense against fraud.

Balancing Safety and User Experience

The paper explores user attitudes toward verification measures and the delicate balance between safety and usability. While users increasingly demand protection from fraud, they also expect seamless, frictionless experiences on dating platforms. Striking this balance is one of the most significant challenges facing the industry.

Research included in the report shows that most users are willing to undergo verification if the process is quick and respectful of privacy. However, overly burdensome checks can drive users away or create barriers to genuine connection. Platforms must design safety measures that feel protective rather than intrusive.

Future Directions for Identity Verification

Additional topics covered in the white paper include the potential for reusable identity signals and stronger age assurance. Reusable identity verification could allow users to prove their authenticity across multiple platforms without repeating the process. This approach could reduce friction while maintaining security standards.

Stronger age assurance measures are also becoming increasingly important as platforms face regulatory pressure to protect minors. The report examines various technologies and approaches for verifying user age without compromising privacy. These solutions must be robust enough to prevent circumvention by determined bad actors.

The paper concludes with practical recommendations for dating platforms looking to strengthen their fraud prevention capabilities. These recommendations are designed to be actionable and scalable across different platform sizes and user bases. Implementation will require investment but is essential for maintaining user trust in an era of increasingly sophisticated AI-enabled fraud.