Arceris AI Security Gateway

Arceris is a self-hosted AI security gateway that lets organizations apply security policies to how employees, applications and developer tools use AI. It works at the network layer as a transparent or explicit HTTPS proxy, covering AI chat services such as ChatGPT, Claude and Gemini in the browser, coding tools such as Claude Code, Cursor and VS Code, and applications that call model APIs through a single unified endpoint.

Arceris inspects prompts, uploaded files (including images via OCR) and model responses. Built-in detectors cover personal data, financial identifiers and secrets such as cloud keys and private keys, which can be masked or blocked. It also detects prompt injection, scores risk across multi-turn sessions, applies topic-based policies and can forward content to an existing DLP system over ICAP.

Administrators can control which users, groups or networks may reach which models, route requests to locally hosted models, set token and spending limits, and export structured audit events to a SIEM. It runs entirely on the customer's infrastructure, including air-gapped environments, and integrates with Active Directory/LDAP, Kerberos and OIDC.