Skip to main content

How to govern ChatGPT/Claude Desktop (enterprise)

This guide describes how enterprise IT can route ChatGPT Desktop and Claude Desktop traffic through Talon using DNS and TLS interception. Every conversation from managed devices is then audited and policy-enforced.

Prerequisites: Enterprise-managed devices, internal CA, DNS control, and Talon deployed on an internal server.

How it works​

  1. Enterprise DNS resolves api.openai.com (and optionally api.anthropic.com) to the Talon server IP.
  2. Talon accepts HTTPS connections using certificates issued by your internal CA for those hostnames.
  3. The gateway pipeline runs (PII scan, policy, evidence); Talon forwards to the real provider using vault-stored API keys.
  4. Employees use ChatGPT Desktop or Claude Desktop as usual; the apps are unaware that Talon is in the path.

Architecture​

[Employee laptop] → DNS: api.openai.com → [Talon server]
                                        → TLS (internal CA)
                                        → Gateway (identify by source IP, PII scan, policy, evidence)
                                        → Forward to real api.openai.com with vault key

Caller identification for desktop apps is by source IP range (e.g. engineering VLAN), since the apps do not send Talon-issued API keys. Configure identify_by: "source_ip" and source_ip_ranges in the gateway callers block.

Configuration (schema)​

The gateway config supports a network_interception block for this use case. Implementation of TLS termination and host-based routing is planned for a later phase; the schema is available now:

gateway:
  # ...
  callers:
    - name: "chatgpt-desktop-engineering"
      identify_by: "source_ip"
      source_ip_ranges: ["10.1.1.0/24"]
      tenant_id: "default"
      team: "engineering"
      policy_overrides:
        max_daily_cost: 15.00
        pii_action: "warn"

  network_interception:
    enabled: false   # Phase 2
    intercept_hosts:
      - original: "api.openai.com"
        provider: "openai"
      - original: "api.anthropic.com"
        provider: "anthropic"
    tls:
      cert_dir: "/etc/talon/certs/"

Steps (when network interception is implemented)​

  1. Deploy Talon on an internal server.
  2. Generate TLS certificates for api.openai.com and api.anthropic.com using your internal CA.
  3. Deploy the internal CA to managed devices (e.g. via MDM/GPO).
  4. Update internal DNS so api.openai.com and api.anthropic.com resolve to the Talon server.
  5. Store real provider API keys in Talon's vault.
  6. Configure the gateway with network_interception enabled and the appropriate callers (source IP ranges and policy overrides).
  7. Verify: an employee uses ChatGPT Desktop or Claude Desktop; Talon logs every request and applies policy.

Compliance outcome​

  • Every ChatGPT/Claude conversation from managed devices is audited.
  • PII in prompts is detected and logged (or blocked/warned per policy).
  • Cost is tracked per team via source IP → caller mapping.
  • Policy violations can be blocked or limited by model and cost.

Current status​

  • Now: Gateway supports caller identification by API key and by source IP (identify_by: "source_ip", source_ip_ranges). Use the gateway as a proxy by pointing clients at http://talon:8080/v1/proxy/openai/v1 (trailing /v1 so paths like chat/completions become .../v1/chat/completions) with a Talon-issued key.
  • Phase 2: Full DNS interception (Talon as TLS endpoint for api.openai.com / api.anthropic.com) so desktop apps need no config change.

You're done​

You now know how to route ChatGPT/Claude Desktop traffic through Talon (DNS + gateway, or proxy with Talon-issued key). Talon logs every request and applies policy by caller or source IP.

Next steps:

I want to…Doc
Cap cost per team or by source IPHow to cap daily spend per team or application
Route a Slack bot or script through TalonAdd Talon to your existing app, Slack bot
Export evidence for auditorsHow to export evidence for auditors
Understand the gateway pipelineWhat Talon does to your request