TLDR: Transmit Security’s ‘Blinded by the Agent’ research reveals that consumer AI agents are increasingly bypassing traditional fraud detection systems, leading to a projected surge in fraud losses and operational workload for businesses. The industry is currently unprepared for this shift, as existing fraud controls designed for human behavior are proving ineffective against AI-driven interactions.
TEL AVIV, Israel & BOSTON – Transmit Security has issued a critical warning regarding the escalating threat posed by consumer AI agents to conventional fraud detection systems. According to their latest ‘Blinded by the Agent’ research, these advanced AI agents are creating significant blind spots for enterprises, which are largely unprepared for this evolving challenge.
The report highlights how AI agents, such as OpenAI’s ChatGPT Agent, operate on behalf of users across various platforms, effectively bypassing established security measures like behavioral biometrics, device fingerprinting, and bot detection. These traditional controls were originally designed to analyze human behavior, making them vulnerable to the sophisticated, automated actions of AI agents.
Mickey Boodaei, CEO and Co-Founder of Transmit Security, emphasized the urgency of the situation, stating, “Fraud controls today were built for a world where humans click the buttons. But now, AI is clicking them for us — and the systems can’t tell the difference between AI operated by legitimate users and AI operated by fraudsters.” He further warned, “If we don’t act now, the rise of agentic AI will break the fraud stack as we know it.”
The research unveils several alarming predictions:
Over 60% of online traffic to retailers is currently attributed to bots, not human users. This figure is projected to exceed 90% in the near future as AI agents become more prevalent.
Fraudsters are increasingly leveraging legitimate AI agents, which effectively blinds core detection layers, making it harder to distinguish between legitimate and malicious activity.
A staggering increase of up to 500% in fraud losses is projected over the next few years due to the breakdown in current detection capabilities.
Existing fraud systems are inadvertently flagging legitimate agent transactions, leading to an increase in false declines and a detrimental impact on customer experience.
Fraud teams are expected to face a 2–3 times greater operational workload over the next 12–18 months simply to maintain current levels of protection.
David Mahdi, CIO of Transmit Security, underscored the broader implications, stating, “This is not just about fraud — it’s about trust. When the AI agent becomes your user’s digital proxy, your systems must adapt. Identity, fraud, and authentication platforms need to be re-architected to recognize and verify intent — not just inputs.”
Also Read:
- Darwinium Unveils Advanced AI Tools to Combat Adversarial Fraud
- OWASP Unveils Comprehensive Security Guidance for Agentic AI Applications
The findings underscore the urgent need for businesses, particularly financial institutions and online merchants, to re-evaluate and re-architect their fraud prevention strategies to incorporate predictive AI systems capable of discerning fraudulent activity within agent-generated interactions.
