Intercis and Protect AI
protectai.com redirects to Palo Alto's Prisma AIRS, so the comparison is with a platform inside a large security vendor.
| Intercis | Protect AI, from their own pagesRead 12 September 2026: protectai.com, paloaltonetworks.com/ai-security/prisma-airs, paloaltonetworks.com/ai-security/ai-gateway, paloaltonetworks.com/legal-notices/trust-center. | |
|---|---|---|
| In-path proxy or SDK | In-path proxy, out of your process. Your base URL points at us and the agent's code does not change. We govern Anthropic's POST /v1/messages and OpenAI's POST /v1/responses and POST /v1/chat/completions. |
protectai.com returned a single 301 to paloaltonetworks.com/ai-security/prisma-airs. Their AI Gateway page reads: "Get secure access to more than 3000 LLMs, MCP servers and tools", and the Prisma AIRS FAQ names deployment shapes such as Network Intercept and API Intercept. |
| What happens to a denied action | The tool_use block is taken out of the model's response and one text block naming the policy takes its place, so the agent runtime has no tool call to run. |
"Secure your agent ecosystem by verifying every agent identity and enforcing real-time security to stop unauthorized actions as you scale from pilot to production." (paloaltonetworks.com/ai-security/prisma-airs) What the caller receives when an action is stopped, their pages we read do not say. |
| What the buyer sees per event | One row per governed tool call: the agent, the tool name, a short excerpt of the target, the verdict, the policy, the wire route, a command fingerprint and the structured tool input. Rows are hash-chained per tenant by a database trigger. | Their AI Gateway page lists Observability among six capability cards. What one event record holds, their pages we read do not say. |
| Published pricing | Yes. $200 a month for the first agent and $190 for each additional agent, on the pilot page. | No. We found no price on the pages we read, and the conversion path is a demo request. |
| Certification | No. Intercis is not SOC 2 certified and no audit is underway. An evidence-package export is a roadmap item without a date. | paloaltonetworks.com/legal-notices/trust-center names no certification on the page itself and carries a "Request a Security Report" button. trust.protectai.com returned a page-not-found. |
| Self-hosted | No, hosted only today; self-host licensing is undecided. | Whether you can run it in your own infrastructure, their pages we read do not say. |
What we do to a denied call, and the five costs
When policy denies a tool call, the proxy does not hand a verdict back to your code: it
replaces the provider's response. The model's tool_use block is gone, one
assistant text block naming the policy is in its place, and stop_reason is
end_turn, so the agent runtime has nothing to execute. That is
_make_blocked_response in apps/proxy/main.py, read at commit
87ccd47 on 12 September 2026.
Being in-path is a setting, not a law
Your base URL points at us and a compromised process can point somewhere else, so the control holds only if you also stop the agent reaching the provider directly.
Only the provider API wire is visible
We judge tool calls that cross the Anthropic and OpenAI API wire. A hosted MCP call has already run by the time the response reaches us, so it is observed, not governed: it writes one observe row and is never judged or blocked.
Streaming is buffered before anything is judged
The proxy collects a streamed response in full before judging it, so your agent waits the length of the stream, and nothing in the repository measures that delay.
We are a new place your prompts pass through
Your prompts and tool inputs cross our proxy in plaintext, the classifier call goes out on an Intercis credential carrying the tool name and the full tool input, and how a self-hosted deployment would be licensed is not decided.
The provider routes are ours to maintain
We speak Anthropic's messages route and OpenAI's responses and chat-completions routes, and another provider is engineering work on our side.