Cohere’s North AI agent platform is designed for private deployment inside an organization’s infrastructure, keeping sensitive data secure behind firewalls. Image Source: ChatGPT-4o
Cohere Launches ‘North’ AI Agent Platform with Private, Secure Deployment for Enterprises
Key Takeaways:
Cohere has introduced North, an AI agent platform designed for private deployment to protect sensitive enterprise and government data.
North can run on on-premise servers, hybrid clouds, VPCs, or air-gapped environments, without Cohere accessing customer data.
The platform is designed to work on as few as two GPUs, with security features like granular access control, continuous red-teaming, and compliance with GDPR, SOC-2, and ISO 27001.
North supports chat and search functions powered by Cohere’s Command and Compass technologies, plus enterprise-focused reasoning capabilities.
Cohere has piloted North with organizations such as RBC, Dell, LG, Ensemble Health Partners, and Palantir.
Cohere Targets Data Security Concerns with North
While AI agents are promoted as a way to reduce repetitive work, many enterprises remain cautious, citing data security risks.
Cohere, a Canadian AI company specializing in large language models for enterprise applications, aims to address those concerns with North, a new AI agent platform that can be deployed entirely inside a customer’s infrastructure, ensuring data remains behind the organization’s own firewall.
“LLMs are only as good as the data they have access to,” said Nick Frosst, co-founder of Cohere, during a product demo. “If we want LLMs to be as useful as possible, they have to access that useful data, and that means they need to be deployed in [the customer’s] environment.”
Runs Anywhere — Even on Two GPUs
Unlike many AI tools hosted on cloud platforms such as Azure or AWS, North can be installed on an organization’s private servers, hybrid systems, or air-gapped setups.
“We can deploy literally on a GPU in a closet that they might have somewhere,” Frosst said, adding that the platform can run on as few as two GPUs.
North also incorporates security protocols such as granular access controls, agent autonomy policies, and third-party security tests. Cohere says it meets international compliance standards, including GDPR, SOC-2, and ISO 27001.
Beyond Security: AI Agents That Work
Cohere has raised $970 million to date, with its most recent valuation at $5.5 billion. The company says North has already been piloted with customers including RBC, Dell, LG, Ensemble Health Partners, and Palantir.
North includes chat and search capabilities that can answer customer queries, summarize meeting transcripts, generate marketing copy, and pull information from both internal sources and the web. All outputs come with citations and reasoning chains to enable verification.
The platform builds on Cohere’s Command generative models and Compass multimodal search stack, using a Command variant optimized for enterprise reasoning.
“It goes beyond just Q&A and gets into doing work for you,” Frosst said. “It can make tables, documents, slideshows, and conduct market research.”
Acquisition and Integrations
In May, Cohere acquired Ottogrid, a Vancouver-based company specializing in automated market research tools for enterprises. That capability now supports North’s asset creation and research features.
North can integrate with common workplace apps like Gmail, Slack, Salesforce, Outlook, and Linear, as well as Model Context Protocol (MCP) servers for industry-specific systems.
“As you build confidence by chatting to the model, there’s like a smooth transition that happens between using this as an augmentation to using it as an automation,” Frosst said.
Q&A: Cohere’s North AI Agent Platform
Q: What is Cohere’s North?
A: North is an AI agent platform for private, secure deployment within enterprise and government infrastructure.
Q: How does North protect data?
A: It runs entirely inside the customer’s environment, ensuring Cohere never sees or stores their data.
Q: What infrastructure can North run on?
A: It works on on-premise servers, hybrid clouds, VPCs, or air-gapped systems, requiring as few as two GPUs.
Q: What features does North offer?
A: Chat, search, asset creation, market research, and integrations with popular workplace tools.
Q: Who has tested North so far?
A: Organizations such as RBC, Dell, LG, Ensemble Health Partners, and Palantir.
What This Means
With North, Cohere is positioning itself as a solution for enterprises that want AI-powered automation but cannot compromise on data security. By enabling full deployment inside the customer’s own infrastructure, it removes one of the biggest adoption barriers for AI in regulated sectors.
North’s combination of enterprise reasoning, auditable outputs, and integration with existing tools could make it appealing to industries ranging from finance and healthcare to government. In a market where many AI platforms are tethered to public clouds, Cohere’s “run it in your own closet” approach could set it apart in the growing AI agent space.
Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.