Device intelligence company FingerprintJS Inc. today launched a preview of two products built to identify traffic from artificial intelligence assistants, addressing a detection gap that has opened as more web requests arrive without a browser.

The first, AI Assistant Detection, gives businesses real-time visibility into traffic from major AI assistants, including OpenAI Group PBC’s ChatGPT, Google LLC’s Gemini and Anthropic PBC’s Claude. It runs on the second product, the Automation Intelligence API, which identifies automated traffic without requiring any client-side JavaScript.

The launch builds on Authorized AI Agent Detection, which Fingerprint released in February to identify autonomous AI agents that take actions on a user’s behalf. The new tools extend that coverage to assistants that browse and summarize content, giving security teams a single view across both categories.

Most web security and analytics tools were built on the assumption that traffic comes from humans opening browsers and JavaScript-based detection worked while that assumption held. AI assistants break it. They reach websites over HTTP to pull content, summarize documentation and conduct research without ever loading a page. Google’s Gemini Spark, announced at I/O 2026, for example, runs on dedicated cloud virtual machines that operate without a user opening a browser.

The blind spot has drawn malicious actors. Scrapers and low-quality bots have found that spoofing the user agent of a popular AI assistant is an easy way past bot defenses, because operators do not want to risk blocking legitimate assistant traffic and losing a new discovery channel.

Fingerprint’s AI Assistant Detection operates at the HTTP level rather than relying on browser signals, letting businesses see which AI tools are accessing their content and flag impersonators before they scrape data or skew analytics. The company verifies assistants through IP-range matching for services such as ChatGPT and Claude and DNS-based verification for Gemini.

The Automation Intelligence API can be deployed at the content delivery network edge, in middleware, or on any backend across cloud platforms. Alongside each classification, it returns IP and network risk signals, including proxy, virtual private network, TOR and geolocation data, so teams can throttle, step up, or allow traffic rather than apply a blanket block.

“The web is going browserless and the pace of that shift is faster than most security stacks were built to handle,” said Valentin Vasilyev, co-founder and chief technology officer at Fingerprint. “The question is no longer ‘Is this a bot or a human?’ It’s ‘Can I trust this visitor, whoever it is?'”

AI Assistant Detection is available in preview to a select group of customers at no additional cost to those using the bot detection Smart Signal.

Fingerprint is a venture capital-backed startup that has raised $77 million in funding, including a round of $32 million in November 2021. Investors in the company include Craft Ventures, Nexus Venture Partners, Uncorrelated Ventures, Hack VC and Entrepreneurs Roundtable Accelerator.

Image: Fingerprint