Artificial intelligence fraud prevention platform Darwinium has launched a new solution to secure agentic commerce. As AI agents increasingly make purchases for consumers, businesses face a new security challenge. Darwinium's agent intent intelligence evaluates behavioral intent and dynamically applies friction based on risk, going beyond simple authentication.
The solution, deployable across major content delivery networks like Cloudflare and AWS CloudFront, distinguishes between verified AI agents, human users, and malicious automation. It analyzes real-time behavior and journey context to permit, verify, challenge, or prevent interactions.
"The real challenge is distinguishing trusted automation from abuse and applying the right level of trust at the right moment," said Chief Executive Alisdair Faulkner. "Our intent intelligence gives organizations the intelligence to do exactly that."
Darwinium's system offers a governance layer for delegated digital activity, evaluating requests in real time with four adaptive responses: permit, verify, challenge, and prevent. Enforcement at the edge ensures controls are applied at critical moments like login or checkout without introducing latency.
The new solution interoperates with evolving payment and identity protocols, including Visa Trusted Agent Protocol and Mastercard Agent Pay. It is available immediately across all Darwinium edge deployments.