Google launches Cloud Fraud Defense, positioning it as reCAPTCHA’s next evolution for the agentic web

Google on Tuesday unveiled Cloud Fraud Defense at its Cloud Next ‘26 conference, pitching the system as a trust platform for the emerging “agentic web” and the next evolution of reCAPTCHA. The agentic web—where autonomous AI agents can reason, plan, and execute complex transactions using open web standards—promises new kinds of digital interactions but also new avenues for abuse.

Google said Fraud Defense is built to verify the legitimacy of humans, bots, and AI agents, supplying businesses with intelligence to secure online interactions and commerce using some of the same global signals that protect its own ecosystem.

As described by the company, Fraud Defense introduces tools to measure and control agentic activity on websites, including a new dashboard for “agentic activity measurement.” Google said it is integrating with industry standards such as Web Bot Auth and SPIFEE, alongside traditional methods, to identify, classify, and analyze agent-driven traffic and to connect agent and human identities for a clearer view of risk and trust.

The company also highlighted policy tools to govern how agentic traffic is handled. reCAPTCHA remains the platform’s core bot-defense pillar. According to Google, existing reCAPTCHA customers become Fraud Defense customers automatically, with no migration required, no action needed, no change to pricing, and no changes to current site keys or integrations.

Google outlined a three-pronged approach. First, it aims to prevent evolving threats with fraud intelligence used across many of its services. As attacks shift from bot automation and invalid traffic to agent takeover and large-scale, AI-driven synthetic identity fraud, the company said Fraud Defense is designed to identify emerging threats before they reach a site.

Google said this visibility is underpinned by a fraud intelligence graph that already protects 50% of Fortune 100 companies and more than 14 million domains globally, arguing that this scale can provide “collective immunity” and verified trust beyond what local data can deliver.

Second, the platform focuses on securing the customer journey end to end, correlating telemetry from registration and login through payment and checkout. Google said this unified trust model can surface complex, multi-stage fraud campaigns that point solutions miss and has demonstrated a 51% average reduction in account takeover.

Third, Google emphasized reducing friction to support growth. Fraud Defense is designed to be invisible for most users, replacing disruptive puzzles with background verification.

The company said its trust model lets businesses precisely block malicious bots, humans, and agents while welcoming legitimate users—including AI shopping assistants that drive a projected 25% increase in average order value, according to the 2025 Shopify Retail Report.

Google framed the launch as part of a broader push to enable a safer agentic web, with reCAPTCHA customers folded in immediately under the expanded platform.