Feedzai Unveils ScamAlert and TRUST Framework, Fraud Prevention with Responsible AI

3 min read
Feedzai Unveils ScamAlert and TRUST Framework, Fraud Prevention with Responsible AI

Feedzai, the global leader in AI-native financial crime prevention, today announced two groundbreaking initiatives at the HumanX conference: ScamAlert, an advanced GenAI-powered fraud prevention agent, and the TRUST Framework, a comprehensive approach to ensuring ethical, scalable, and responsible AI deployment. These innovations reinforce Feedzai’s dual mission: stopping financial crime while ensuring AI is used responsibly.

Scams have escalated into a global crisis, costing consumers over $1 trillion annually as fraudsters leverage GenAI to execute increasingly sophisticated attacks. Traditional fraud prevention tools often struggle to address these evolving threats, particularly as many scams involve customers unknowingly validating fraudulent transactions.

ScamAlert is a revolutionary AI agent designed to break the scam cycle at its source. By providing real-time alerts and actionable insights, ScamAlert transforms customers from passive victims into active defenders against fraud. Banks can integrate ScamAlert in multiple ways, from self-service payment verification to risk engine augmentation, allowing for real-time scam detection before transactions are processed.

Related startups

“Scammers are evolving fast, using AI to trick even the most cautious consumers,” said Nuno Sebastiao, CEO and Co-Founder of Feedzai. “With ScamAlert, we’re giving people a powerful AI assistant that warns them in real time—because the best fraud prevention starts before money leaves your account. Fraud prevention is not just a technology problem; it requires collaboration between banks, tech companies, and consumers.”

ScamAlert’s multimodal AI analyzes text, images, and transaction data to detect scams before they happen. Customers can upload screenshots of suspicious emails, text messages, marketplace listings, or invoices directly through their mobile banking app, receiving instant feedback on the legitimacy of their transactions. By leveraging Feedzai’s global intelligence network, ScamAlert enhances banks’ risk assessment models, ensuring that fraud detection remains proactive and effective.

Attendees at the HumanX conference had the opportunity to engage with ScamAlert through an interactive game, testing their scam detection skills against the AI agent in real-time.

Alongside ScamAlert, Feedzai introduced the TRUST Framework, a five-pillar approach designed to embed responsible AI principles into every stage of GenAI development. As AI adoption accelerates across industries, businesses must ensure that their models are transparent, robust, unbiased, secure, and thoroughly tested. The TRUST Framework provides a structured approach to achieving these goals, ensuring that AI remains ethical, explainable, and compliant with evolving regulations.

“AI is becoming increasingly commoditized, but differentiation comes from responsibility and ethics,” said Pedro Bizarro, Ph.D., Co-Founder and Chief Science Officer at Feedzai. “The TRUST Framework ensures that organizations build AI systems that are fair, explainable, and secure—just as the automotive industry shifted from valuing speed to prioritizing safety and sustainability, we must design AI with the same foresight. By combining cutting-edge fraud detection with responsible AI practices, Feedzai is taking a leadership role in shaping the future of AI-driven financial security.”

Together, ScamAlert and the TRUST Framework solidify Feedzai’s commitment to a dual mission: stopping financial crime while ensuring AI is used responsibly.

Sponsored content disclosure: This article contains sponsored content. Our editorial standards remain paramount — opinions, analysis, and conclusions are independent and were not dictated by the sponsor. We accept compensation for distribution and promotion, never for editorial direction. See our partner program for how sponsorships work.

© 2025 StartupHub.ai. All rights reserved. Do not enter, scrape, copy, reproduce, or republish this article in whole or in part. Use as input to AI training, fine-tuning, retrieval-augmented generation, or any machine-learning system is prohibited without written license. Substantially-similar derivative works will be pursued to the fullest extent of applicable copyright, database, and computer-misuse laws. See our terms.