Amanda | ORGN

Today we're launching Blackbit ZERO - the agentic harness for security teams. An autonomous environment that brings cyber-native models, agent swarms and tools under your command to test, trace and verify your own systems.

The calm before the storm

ORGN has removed models from Anthropic, OpenAI, SpaceX/xAI, Google, Meta, Microsoft and NVIDIA Protecting the privacy and integrity of user data comes first. Only open-source models and models we trust to maintain zero retention remain available. Powerful AI should not make confidentiality optional.

A founder using AI to code faster doesn't want their IP as training data. Powerful AI should not make confidentiality optional ORGN makes confidentiality the standard.

Banks using AI to flag suspicious transactions cannot afford customer data falling into the wrong hands Powerful AI should not make confidentiality optional ORGN makes confidentiality the standard

Unpublished work is research, not training data. AI is for building on discoveries rather than taking credit for them. Helpful AI should not make confidentiality optional, hiding inside convoluted settings. ORGN makes confidentiality the standard.

Investment memos hold your next moves. They do not belong in just any AI tool. Helpful AI should not make confidentiality optional. ORGN makes confidentiality the standard.

Do this once: list AI providers recently caught accessing, retaining, or training on user data when they shouldn't OpenAI Meta Anthropic Google xAI Then decide who you trust with sensitive work.

Before you give AI access to your apps, data, or code, ask: Where is it processed? Who can access it? What is retained? A privacy policy and pledge not to look are not always enough. Sometimes, for sensitive work, you need a system that can't look. That, is ORGN.

We’ve made GLM-5.3 Flash on confidential computing hardware available through the ORGN Gateway. Your prompts and sensitive data can now run encrypted in hardware-isolated environments, with attestation records you can verify. For workloads where a privacy policy is not enough.

Every system has a way in. Find yours first.

Three model types: Abliterated Private Cyber-native Because authorised security work shouldn't trigger arbitrary refusals Because prompts and findings should stay confidential Because defenders should have access to intelligence built explicitly for their line of work

ops @orgn_official is cooking check out blackbit.sh

And that's why we built ORGN on confidential computing The point is not simply to promise privacy, but create an architecture where access is technically constrained We effectively built a vault around customer workloads, then made sure to lock everyone, and even ourselves, out Proud to be building this as part of the #NVIDIAInception program.

Fable and GPT-6 will refuse your prompts. CyberGLM won’t. Abliterated models are now on the ORGN Gateway, private, with zero data retention For when you need fewer refusals and don’t want your prompts retained.

ORGN Scan is our public trust layer. It lets anyone inspect attestation records from our Sandboxes and TEE model requests. You can verify that data was processed in isolation, encrypted, and with the expected software state. If we make a privacy claim, we show the evidence behind it

Cursor is for devs who want AI in a familiar editor. Claude Code is for devs who want agentic coding power. ORGN is for devs who want both, but without handing their code and data to model providers, third parties, or infrastructure they cannot verify.

Here's the thing: OpenAI has "privacy policies", but they don't make accessing user data impossible, when that is exactly what is needed for real privacy. We actually built ORGN around that premise: making access to our user's cloud session data impossible, even for us.

DeepSeek-V4.1 Flash is now available on the ORGN Gateway with zero data retention. Your prompts aren’t kept. No training on your data. We’re also working to deploy it on confidential hardware, with verifiable privacy protections built down to the silicon. Stay tuned 🖤

We dogfooded our own stack to build ORGN CRM: an encrypted, private CRM. CRM's contain some of a company's most sensitive information: contact details, relationship history, and often, confidential email messages. Privacy was thus non-negotiable, so we built it using our confidential IDE, Intel TDX Sandboxes, and AI models deployed on confidential computing chips. ORGN CRM is designed to keep sensitive data out of platforms and AI models that may retain it, make it retrievable, or use it for training. We’re considering making it available more broadly. Should we?