ShieldLabs
Start Free

AI Agent Detection Solution

AI agent detection.
Know who is visiting.

AI agents and ordinary bots need different handling. ShieldLabs identifies AI bots, AI browsers and their providers, helping you protect access, recognize useful automation and distinguish good bots from bad bots.

Start Free

Identify the client

Recognize the AI bot reaching your product.

See the provider

Identify traffic from OpenAI, Anthropic and other supported providers.

Distinguish good bots from bad

Welcome useful automation and identify abusive bot activity.

Understand the impact

Get AI client detections with ready-made risk and traffic quality results.

AI client classification

Recognize the type.
See the provider.

See which AI client is visiting and who operates it. The current coverage separates AI bots, search and web crawlers, and AI browsers.

AI bots

7 clients
  • OpenAIGPTBot, ChatGPT-User
  • AnthropicClaudeBot, Claude-User
  • PerplexityPerplexity-User
  • Mistral AIMistralAI-User
  • DuckDuckGoDuckAssistBot

Search and web crawlers

31 clients
  • OpenAIOAI-SearchBot
  • AnthropicClaude-SearchBot, anthropic-ai, Claude-Web
  • PerplexityPerplexityBot
  • BaiduBaiduspider
  • NaverYeti
  • GoogleGoogleOther, Googlebot, Storebot-Google
  • Microsoftbingbot
  • AppleApplebot
  • YandexYandexBot
  • MetaFacebookBot, meta-externalagent, Webindexer
  • HubSpotHubSpot Crawler
  • CohereCohere crawler
  • AmazonAmazonbot
  • ByteDanceBytespider
  • Webz.ioOmgilibot, Omgili
  • Huawei (Petal Search)PetalBot
  • TimpiTimpibot
  • Shenma SearchYisouSpider
  • Dataprovider.comDataprovider
  • Headlineev-crawler
  • Diffbot
  • You.comYouBot
  • Common CrawlCCBot
  • SE RankingSeRanking

AI browsers

2 clients
  • Browserbase
  • CloudflareCloudflare Browser Run
Explore browser bot detection

Agent traffic handling

Tell good bots from bad.
Respond to risky activity.

ShieldLabs helps distinguish good bots from bad bots by bringing client identity and risk evidence together. Welcome useful crawlers and AI assistants, and restrict automation behind fake signups, repeated claims and abusive access.

Allow

Welcome authorized agents for useful browsing and data retrieval.

Enrich

Add provider, identity and risk context to agent requests.

Throttle

Manage request volume and access for multi-step workflows.

Block

Restrict impersonation, scraping and abusive access.

Identity and fraud prevention

Identify AI traffic.
Prevent AI-driven abuse.

A familiar AI name can accompany risky requests. ShieldLabs detects browser automation and account abuse to help stop fake registrations, automated submissions and promo abuse.

ShieldLabs MCP

Investigate traffic with your AI assistant.

Connect ChatGPT, Claude or another MCP client to ShieldLabs. Review identification history, understand risk signals and summarize activity for an account or device.

Connect through MCP

Why was this visit marked as suspicious? Explain the detected risk signals.

Understand risk
Start Free

The ShieldLabs platform

AI identity.
The risk context around it.

Bring AI client detection together with identification, device and network intelligence, risk scoring and traffic quality analytics.

Enterprise-level functionality. Without enterprise pricing.

Accurate identification

Recognize returning visitors even when they clear cookies, use incognito mode or change their IP address. Detect linked users with different email addresses.

99.9%

Identification accuracy

UsersDevicesVisitorsPublic IPsLocal IPsCookiesSessions

High-Risk Event detection

Detect account abuse out of the box by automatically identifying linked users. No fraud model training is required.

Risk signals

Get ready-made detections of automation, masking and spoofing, with named risk signals behind each result.

VPNProxyPrivate RelayTor NetworkCheck IncompleteAnti-Detect BrowserBanned IPIncognito modeGood botTimezone MismatchBad bot

Bot & AI traffic detection

Detect bot activity from browser automation signals. Recognize useful crawler and AI traffic alongside abuse detections.

Risk scoring

Score every visit from 0 to 100 and classify it as trusted, suspicious or dangerous. See the signals behind each score.

Device fingerprinting

Recognize devices by their fingerprints. Identify the browser, operating system and device type. Detect headless browsers and tampering.

IP & Network Intelligence

Reveal the real IP address and geolocation behind masking. Detect VPNs, proxies and private relays.

Traffic quality scoring and risk analytics

Get one score for all your traffic and ready-made quality results by source, channel, referrer and UTM campaign.

13Trusted
0100
Trusted61.0%Suspicious24.5%Dangerous14.5%
Start Free

Analytics dashboard

See agentic traffic.
Understand its impact.

Get AI client classifications, available providers and ready-made risk assessments alongside linked users and devices. Find the sources bringing useful AI traffic and detected abuse.

Start Free

How it works

Start protecting your product.

Connect once and start receiving ready-made fraud and abuse detections.

  1. 01

    Add the snippet or SDK

    Connect your product and start identifying visitors and users.

  2. 02

    Receive ready-made detections

    Get detected fraud and account abuse in the dashboard or through API integrations.

  3. 03

    Prevent fraud and abuse

    Protect customer accounts, paid access and signup incentives in your product.

Developers

Integrate in
5 minutes.

A single API to identify, detect and score. Receive risk data through API responses and webhooks. Integrate selected data points into your business logic to prevent fraud and abuse.

Start Free
identification.scored
JavaScriptReactNext.jsVueAngularSvelteNode.jsPythonGoPHPJava.NETWordPressTildaShopify

AI Agent Detection

Recognize the agent.
Keep risk in view.

Start Free

FAQ

Understand AI traffic.
Use the identity evidence.

What is AI agent detection?

AI agent detection identifies AI clients in website traffic and adds their client category and provider context. ShieldLabs distinguishes supported AI bots from search and web crawlers, alongside risk and traffic quality results.

How does ShieldLabs help prevent AI-driven abuse?

ShieldLabs identifies AI client context alongside browser automation, risk signals and account abuse detections. Handle useful agents while protecting your product from risky automated activity.

How is it different from browser bot detection?

Bot detection identifies browser automation and other bot activity. AI agent detection adds named AI clients and provider context, so AI traffic can be understood alongside other automation.

Can I analyze AI traffic by source?

AI traffic context can be reviewed alongside visits, devices, users, sources and campaigns, so you can understand its impact on traffic quality.

Which AI bots does ShieldLabs detect?

The current AI bot coverage includes GPTBot and ChatGPT-User from OpenAI, ClaudeBot and Claude-User from Anthropic, Perplexity-User, MistralAI-User and DuckAssistBot. Search and web crawlers are listed separately.

Which AI browsers are supported?

Browserbase and Cloudflare Browser Run are recognized when signature verification succeeds.

Does an AI client name mean the request is trusted?

A client name describes the traffic. It does not authorize access to an account, paid content or a product action. Review identity verification and risk evidence together with the access requirements of your product.

How can I investigate traffic with an AI assistant?

Connect an MCP client to ShieldLabs to retrieve identifications, explain Risk Scores and summarize recent history for an account or device. Summaries describe the records returned by the query.

Can I distinguish useful AI traffic from abusive activity?

Yes. ShieldLabs identifies supported AI bots and crawlers alongside risk results and browser automation detections. Use client identity and risk evidence together to welcome useful traffic and protect your product from abusive activity.