Skip to content
Home

/

Glossary

/

Browser

/

HUMAN Security (PerimeterX)

Browser

5 min read

Also: PerimeterX, PX, Press & Hold challenge, Press and hold captcha, HUMAN Bot Defender, HUMAN Challenge, _px3 cookie, HUMAN Security

HUMAN Security (PerimeterX)

Quick answer

PerimeterX, now HUMAN Security's Bot Defender, is a bot-protection service known for its Press & Hold challenge and for cookies starting with _px, such as _px3. If a scraper or automated browser keeps getting the challenge or a 403, read why anti-bot systems block scrapers and what actually helps.

PerimeterX, now HUMAN Security's Bot Defender after a 2022 merger, is a bot-protection service that scores every request and challenges suspicious visitors with its 'Press & Hold' button, the HUMAN Challenge.

Anti-bot system
Official site ↗

Last verified

What people do with HUMAN Security (PerimeterX)

Key facts

OwnerHUMAN Security (merged with PerimeterX in July 2022; HUMAN was formerly White Ops)
Product name todayHUMAN Bot Defender, part of the HUMAN Defense Platform
ChallengeHUMAN Challenge, a 'Press & Hold' button that also runs proof-of-work checks on the device
Telltale signsCookies prefixed _px (e.g. _px3, _pxhd, _pxvid); 'Please verify you are a human' block page
DeploymentClient-side JavaScript sensor or mobile SDK plus a server-side enforcer at the edge
Pricing modelEnterprise, quote-based

What is HUMAN Security (PerimeterX)?

PerimeterX was a bot-protection company best known for its "Press & Hold" challenge: instead of a picture puzzle, the page asks you to press and hold a button until it fills up. In July 2022 PerimeterX merged with HUMAN Security (the ad-fraud and bot-detection company formerly called White Ops), and the combined company operates under the HUMAN name. The former PerimeterX product is now sold as HUMAN's Bot Defender, part of the HUMAN Defense Platform, alongside account-takeover, fraud and client-side code protection products.

People still search for "PerimeterX" far more than for "HUMAN Bot Defender", partly because many of the technical traces the product leaves (cookies, script paths, block pages) kept their original PerimeterX names after the merger. You will find it on large retail, travel, real-estate and ticketing sites, typically protecting the whole site rather than one form.

How it works

Bot Defender combines a client-side sensor with server-side enforcement:

  • A JavaScript sensor runs in the visitor's browser (or an SDK in a mobile app) and collects signals about the device, browser and behavior, such as how the pointer moves and how the page is interacted with.
  • An enforcer at the site's edge (a CDN integration, web server module or middleware) checks every request against HUMAN's decision engine using those signals, the request itself, and reputation data HUMAN gathers across its customer base.
  • The enforcer acts. Good traffic passes. Suspicious traffic gets a challenge page. Clearly automated traffic is blocked, usually with an HTTP 403.
  • The challenge is the HUMAN Challenge, the press-and-hold button, which HUMAN describes as the default verification for Bot Defender. HUMAN's documentation says that while the visitor holds the button, the challenge runs proof-of-work tests on the device and collects additional signals. Passing it updates the session's cookies so the visitor can continue browsing. Cookies with names beginning _px (for example _px3, _pxhd and _pxvid) are the usual sign that a site uses this product; the names were kept for compatibility after the rebrand. Block pages typically read "Please verify you are a human" or "Access to this page has been denied" and include a reference ID.

    Why it appears in automation

    HUMAN's customers deploy Bot Defender specifically to stop scraping, scalping, credential stuffing and fake account creation, so automated access is what it is designed to catch:

  • Instant 403s for HTTP clients. A plain script (Python requests, curl, a no-code HTTP step, Google Sheets IMPORTXML) never runs the sensor, so it has no valid session.
  • Press & Hold that never completes, or repeats. Automated browsers built with Selenium WebDriver, Puppeteer or headless Chrome expose automation traits, and a scripted mouse-down is not the same as a human hold.
  • Blocked after a few pages. Navigation speed, request volume and IP reputation all feed the decision; see rate limiting. Datacenter IP addresses start with a low reputation.
  • Real users blocked too. VPNs, privacy extensions and unusual browsers can trigger the challenge for ordinary visitors, who usually clear it by completing the hold.
  • The honest options

    A site that runs HUMAN Bot Defender has chosen to keep automated clients out. If you need its data or need to act on it at scale:

  • Look for an official API, data feed, or partner program first.
  • Ask the site to allow-list your use case; enterprise bot-management tools all support allow-lists for known partners.
  • Reduce the footprint: fewer pages, lower frequency, signed-in access to your own account where appropriate.
  • Let a person handle the challenge for low-volume tasks on accounts you own.
  • Read the site's terms and robots.txt, and see is web scraping legal? for the general picture. This page deliberately does not describe ways to evade the product.

    HUMAN, PerimeterX and Autonoly

    Autonoly's agent drives a real, headed Chromium browser, so pages run their JavaScript as they would for you. That does not make it undetectable, and Autonoly does not solve the HUMAN Challenge or use proxies or IP rotation. When a Press & Hold page or a HUMAN block page appears, the agent pauses and gives you the browser through live browser control. You can complete the hold yourself and hand control back. If a site blocks repeatedly, use an official data source rather than retrying.

    Why It Matters

    Many large retail, travel and real-estate sites use HUMAN Bot Defender. Recognizing the Press & Hold page and the _px cookies tells you a failed run is a deliberate bot block, not a selector problem, and points you to an API, permission or a human step instead of endless retries.

    How Autonoly Solves It

    Autonoly does not solve the HUMAN Challenge and has no proxies or IP rotation. Its agent uses a real headed Chromium browser; when a Press & Hold or block page appears, it pauses and hands you the live browser so you can complete it or stop.

    Learn more

    Examples

    • A listing-monitor script receives 403 responses with a _pxhd cookie and a 'Please verify you are a human' page on every request

    • An agent browsing a retail site meets a Press & Hold button; it hands control to the user, who completes the hold and returns control

    • A team that kept hitting HUMAN blocks signs up for the retailer's partner data feed instead of scraping

    Frequently Asked Questions

    Yes, in practice. PerimeterX merged with HUMAN Security in July 2022 and the combined company uses the HUMAN name. The former PerimeterX bot-protection product is now HUMAN Bot Defender.

    It is the HUMAN Challenge, the default verification for HUMAN Bot Defender. Instead of a picture puzzle, you press and hold a button. HUMAN says that during the hold it runs proof-of-work tests on the device and collects extra signals to decide whether you are human.

    Look for cookies whose names start with _px, such as _px3, _pxhd or _pxvid, a block page reading 'Please verify you are a human' or 'Access to this page has been denied' with a reference ID, and a Press & Hold button.

    VPNs, privacy or ad-blocking extensions, unusual browsers and shared IP addresses can lower your score. Completing the hold usually clears it; disabling the VPN or extension often stops it recurring. If you stay blocked, contact the site with the reference ID.

    The site has deliberately chosen to block automated clients, so evading it raises contractual and possibly legal risk. Use an official API or data feed, ask for access, or keep the task small enough that a person can handle any challenge.

    No. Autonoly does not solve the HUMAN Challenge and does not use proxies. When one appears, the agent hands you the live browser so you can complete it yourself.

    Related terms, automations and guides

    Where this concept shows up in practice.

    DefinitionCAPTCHACAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart) is a challenge-response mechanism designed to distinguish human users from automated bots on websites.DefinitionDataDomeDataDome is a commercial bot-protection service that scores every request to a website, app or API in real time and blocks or challenges automated traffic, using an invisible Device Check and a slider CAPTCHA served from captcha-delivery.com.DefinitionCloudflare Bot ManagementCloudflare Bot Management is the set of Cloudflare features (Bot Fight Mode, Super Bot Fight Mode and the Enterprise bot score) that decide whether a request is automated, and answer with a 'Just a moment...' challenge or an error such as 1020.DefinitionHeadless ChromeHeadless Chrome is Google Chrome running without a visible window. It renders pages and runs JavaScript like normal Chrome and is controlled by code, which makes it the standard engine for automated testing, scraping, screenshots and PDF generation.DefinitionRate LimitingRate limiting is a technique that controls the number of requests a client can make to a server within a given time window, preventing abuse and ensuring fair resource distribution.Definitionrobots.txtrobots.txt is the plain-text file at a site's root that tells crawlers which paths they may fetch. It follows the Robots Exclusion Protocol, standardised as RFC 9309 in 2022, and is voluntary: well-behaved bots obey it, others ignore it.GuideAutonoly vs Bardeen: AI Browser Automation Compared (2026)An honest, in-depth comparison of Autonoly and Bardeen for AI browser automation. We compare architecture, features, pricing, and use cases to help you choose the right tool for your automation needs.GuideWeb Scraping Best Practices: Avoiding Blocks, Bans, and Legal IssuesA comprehensive guide to web scraping best practices. Learn how to avoid IP blocks, deal with CAPTCHAs and anti-bot systems, respect legal boundaries, and where AI browser agents help (and where they don't).

    Stop reading about automation.

    Start automating.

    Describe what you need in plain English. Autonoly's AI agent builds and runs the automation for you — no code required.

    See Features