Cohere’s new AI agent platform, North, promises to keep enterprise data secure

AI agent tools promise to reduce drudgery in daily workflows, but many organizations hesitate to adopt them due to data security concerns. Enterprises with trade secrets, companies in regulated industries, and government agencies worry that their data—or worse, their customers’ data—could be compromised or used to train foundation models. Canadian AI firm Cohere aims to address these concerns with its new AI agent platform, North. This platform enables private deployment, allowing enterprises and governments to keep their data safe behind their own firewalls. “LLMs are only as good as the data they have access to,” said Nick Frosst, co-founder and CEO of Cohere, during a demo of North. “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.” Instead of relying on enterprise cloud platforms like Azure or AWS, Cohere can install North on an organization’s private infrastructure so it never sees or interacts with customer data. North can run on on-premise infrastructure, hybrid clouds, VPCs, or even air-gapped environments. “We can deploy literally on a GPU in a closet that they might have somewhere,” Frosst explained, adding that North was designed to run on as few as two GPUs. Cohere also incorporates security protocols such as granular access control, agent autonomy policies, continuous red-teaming, and third-party security tests. North complies with international standards including GDPR, SOC-2, and ISO 27001.

More than private deployments

Cohere, which has raised $970 million and is valued at $5.5 billion, has piloted North with customers including RBC, Dell, LG, Ensemble Health Partners, and Palantir. North’s core features include chat and search capabilities that help users answer customer support inquiries, summarize meeting transcripts, write marketing copy, and access information from internal resources and the web. All responses include citations and “reasoning” chains of thought for auditing and verification. The chat and search functions are powered by Cohere’s existing technologies like Command (a family of generative AI models) and Compass (a multimodal search tech stack). North uses a variant of the Command model trained specifically for enterprise reasoning. “It goes beyond just Q&A and gets into doing work for you. So, [North] has a bunch of asset creation. It can make tables, documents, slideshows, and conduct market research,” Frosst said. In May, Cohere acquired Ottogrid, a Vancouver-based platform that automates high-level market research, further enhancing North’s capabilities. Like other AI agent platforms, North integrates with workplace tools such as Gmail, Slack, Salesforce, Outlook, and Linear. It also supports Model Context Protocol (MCP) servers to connect with industry-specific or in-house applications. “As you build confidence by chatting to the model, there’s a smooth transition from using it as augmentation to using it as automation,” Frosst added.

FAQ

What is Cohere's North platform?

North is Cohere's new AI agent platform designed for enterprise and government use, focusing on private deployment to ensure data security and privacy.

How does North ensure enterprise data security?

North ensures data security through private deployment options, allowing organizations to keep their data behind their own firewalls. It can run on on-premise infrastructure, hybrid clouds, VPCs, or even air-gapped environments, and does not interact with customer data. It also incorporates security protocols like granular access control and agent autonomy policies.

What are the key features of North?

Key features include chat and search capabilities for tasks like answering customer inquiries, summarizing meetings, and writing content. It also offers asset creation capabilities, such as generating tables, documents, and slideshows, and conducts market research. Responses come with citations and reasoning chains for auditing.

What security standards does North comply with?

North complies with international standards such as GDPR, SOC-2, and ISO 27001. It also implements security protocols like granular access control, agent autonomy policies, continuous red-teaming, and third-party security tests.

What kind of infrastructure can North be deployed on?

North can be deployed on various private infrastructure options including on-premise servers, hybrid clouds, Virtual Private Clouds (VPCs), and even air-gapped environments. It is designed to be lightweight, potentially running on as few as two GPUs.

Crypto Market AI's Take

Cohere's North platform directly addresses a critical bottleneck in enterprise AI adoption: data security and privacy. By offering private, on-premise, or air-gapped deployment options, North allows organizations to leverage powerful AI agent capabilities without exposing sensitive data. This aligns with the growing trend of enterprises seeking greater control over their data, especially in industries with strict regulatory requirements. Our platform, AI Agents, explores the evolving landscape of AI agents across various sectors, including finance and trading. The ability of platforms like North to integrate with existing enterprise tools, as mentioned in the article, is crucial for seamless adoption and unlocking productivity gains. This move by Cohere highlights the increasing sophistication of AI solutions tailored for the enterprise market, a trend we continue to monitor closely in our AI Crypto Market Platform.

More to Read:

• AI Agents: Capabilities, Risks, and the Growing Role
• AI-Driven Crypto Scams Surge 456%: Experts Warn No One Is Safe
• Walmart Bets on AI Super Agents to Boost E-commerce Growth

Source: Originally published at TechCrunch on August 6, 2025.