Resemble AI Launches Agent Detection to Expose AI Agents That Appear Human

Behavioral model classifies sessions as person or AI agent; found 98.8% of disguised agents in internal testing vs 3.1%–14.8% for leading bot detection tools

MOUNTAIN VIEW, CA, UNITED STATES, October 29, 2026 /EINPresswire.com/ — Resemble AI, the AI detection models company, today announced the release of Resemble Agent Detection (RAD-S1), a System One model that identifies agentic traffic on websites and in products. The model classifies each session as a person, bot, or an AI agent based on how the visitor behaves and interacts, rather than on what the visitor claims to be, giving enterprises a way to identify, measure and govern agent traffic.

The launch comes a week after JPMorgan Chase, Morgan Stanley, Citigroup, Bank of America and Charles Schwab moved to block AI agents such as Meta’s Muse and OpenAI’s Dots from customer accounts, citing unresolved questions about liability when an agent acts on a customer’s behalf.

Their caution follows a string of other incidents this year showing what agents are now capable of. In September, Gambit Security reported that a single operator used AI agents to compromise at least 27 companies and steal more than 600,000 credit card records. Weeks later, seven South Korean financial institutions disclosed agent-related breaches. And in Australia, an OpenAI agent bypassed access controls on a government Medicare statistics portal during an internal evaluation, an incident that went unreported for months.

AI Agents Don’t Behave Like Bots, but Most Security Tools Still Can’t Tell Them Apart
Traditional bots announced themselves or followed fixed scripts that left recognizable fingerprints. AI agents are designed to complete multi-step tasks and work around obstacles, usually inside a real browser or a customer’s own signed-in session, and often without identifying themselves. Existing bot detection tools were built to recognize the traditional fingerprints and not to categorize visitors by behavior.

Security leaders now face three crucial decisions: how to distinguish people from AI agents, where to route human and agent traffic based on what each is there to do, and how to stop malicious agents without adding friction to checkout and transactional paths. Personal agents promise consumers faster checkouts and simpler account management, but that promise is currently overshadowed by the security infrastructure needed to safely support it.

Detecting the Behavior of an Agent
Resemble Agent Detection addresses the first of those decisions. For every session, it analyzes cursor movement, typing rhythm and on-page actions, alongside browser and environment signals, and returns a result of Person, AI Agent or Bot. Because the result is based on behavior, the product can identify agents that run in real browsers and even disguise their identity.

The product runs on RAD-S1, a behavioral model developed in-house by Resemble AI. It does not block or route traffic on its own; customers decide how to act on each verdict, whether by allowing an agent, redirecting it to an API or MCP server, restricting it from sensitive areas or stopping destructive actions.
In internal testing across eight website stacks, RAD-S1 identified at least 98.8% of agents built to conceal their automation, including agents running on stealth browsers and agents that took over a clean browser session. Six widely used bot detection tools tested on the same visits identified between 3.1% and 14.8%.

“We already have a clear workflow for humans shopping and banking online: confirm who the person is, check what they’re allowed to do, then give them access to match,” said Zohaib Ahmed, CEO and co-founder of Resemble AI. “Agents will inevitably need to follow the same rules, but that’s an impossible task unless you can tell what’s an agent and what’s a person. Detection is the first step, and the companies that can securely embrace authorized agent traffic will win the business those agents bring.”

Key Capabilities
– Visibility into agent traffic. Every session is classified as Person, AI Agent or Bot, so security teams can measure how much traffic comes from agents and which pages and flows they reach.
– Behavior-based detection. Verdicts combine cursor movement, typing rhythm and actions with browser and environment signals. A spoofed user agent or disguised session does not change the result.
– Customer-controlled policy. Agent Detection does not proactively block users or sessions. Teams set responses page by page and can pair verdicts with existing identity systems or emerging standards such as Sierra’s Personal Agent Protocol.
– No customer friction. Visitors encounter no puzzles or challenges. The script loads asynchronously, and doesn’t collect typed text or passwords.

Start Detecting Agents Today, Free
Companies do not need to wait for an industry standard or a procurement cycle to see how much of their traffic is agentic. Resemble Agent Detection is available now: simply create a free Resemble AI account, add a single script tag, and the first 1 million sessions are classified at no cost. Beyond that, pricing is $1 per 1,000 sessions. Get started at app.resemble.ai.

About Resemble AI
Resemble AI is an AI detection models company that delivers real-time, accurate detection results combined with intelligence and next-action recommendations. It’s what enterprise teams need to detect a threat, protect their content, and make informed decisions quickly. Our models work across every modality, integrate with your workflows, and adapt to your deployment needs to make detection comprehensive, seamless, and secure. Learn more at resemble.ai.

Clint Bagley
Resemble AI
clint.bagley@resemble.ai
Visit us on social media:
LinkedIn
Instagram
YouTube
X
Other

Resemble AI – Agent Detection (RAD-S1) – Explainer Video

Media gallery