Calabi Labs · Guide · 2026-06-07
Fingerprintjs unveils ai assistant detection to filter chatgpt gemini
FingerprintJS AI Assistant Detection: Filter ChatGPT, Gemini & Claude Traffic in Real Time
FingerprintJS has launched AI Assistant Detection — a new capability that identifies and filters traffic from AI assistants like ChatGPT, Google Gemini, and Anthropic Claude directly within your server infrastructure. It runs on the Automation Intelligence API and works at the HTTP level, with no JavaScript required.
What It Does
AI Assistant Detection gives website and API operators real-time visibility into which visitors are AI assistants — and which are spoofed bots pretending to be them. It distinguishes authentic AI assistant traffic from forged traffic by checking server-side signals that are hard to fake.
How It Works
The system uses two primary verification methods:
- IP-range matching — Validates that traffic originates from known IP ranges used by ChatGPT and Claude's infrastructure
- DNS-based verification — Confirms Gemini traffic by resolving the requesting domain against known AI assistant DNS records
This server-side approach means detection happens before any request reaches your application, letting you block, flag, or route AI traffic however you choose.
Why It Matters Now
AI assistants are increasingly crawling the web to train models, summarize content, and power agentic workflows. Most bot detection tools weren't built for this — they look for browser automation signatures, not AI-specific traffic patterns. Fingerprint's new tool addresses a gap that's widening fast:
- Spoofed AI traffic — Anyone can set a user-agent string to "ChatGPT" and bypass naive bot filters
- Content scraping at scale — AI agents pull content without respecting crawl budgets or terms of service
- Analytics pollution — AI visits inflate your metrics and skew audience data
- Security blind spots — Traditional fingerprinting misses browserless, headless, and agentic traffic
Key Features
- Real-time identification of ChatGPT, Gemini, and Claude traffic
- Distinguishes genuine AI assistant visits from spoofed traffic
- HTTP-level detection — no client-side JavaScript needed
- Works alongside existing bot protection or as a standalone signal
- Integrated into the Automation Intelligence API for broader AI traffic visibility
Use Cases
- Block AI training crawlers that ignore robots.txt
- Filter AI traffic from analytics to get cleaner human visitor data
- Control API access by identifying AI agents making automated requests
- Enforce content licensing by detecting when AI services republish your material
Availability
AI Assistant Detection is currently in preview. Fingerprint has also launched a broader Automation Intelligence API that covers the full spectrum of AI agent and assistant traffic beyond just the three major chatbots.
The bottom line: FingerprintJS now lets you see, filter, and act on AI assistant traffic at the server level — catching both legitimate AI crawlers and spoofed bots pretending to be them.
Try Calabi free at calabilabs.com — 10 cleans, no card.