ViqusViqus
Navigate
Company
Blog
About Us
Contact
System Status
Enter Viqus Hub

Cohere Unveils North 2: Enterprise AI Agent Platform with Enhanced Control and Orchestration

AI Agents Enterprise AI Token Governance Orchestration On-Premises AI Language Models
October 05, 2026

This summary and analysis were generated by AI from the original article at AI – SiliconANGLE and may contain errors (how Viqus works). Read the source for full details.

Viqus Verdict Logo Viqus Verdict Logo 7
Operationalizing Agency
Media Hype 6/10
Real Impact 7/10

Article Summary

Cohere has released North 2, a major upgrade to its platform designed for running sophisticated AI agents within large organizations. The update directly addresses the historical trade-off between security, capability, and cost control that enterprises face when deploying agents. Key features include a redesigned orchestration system for multi-step tasks, the ability to share and reuse agents across teams, and a new 'skills' feature for packaging reusable logic. Crucially, North 2 introduces granular token spending controls, allowing administrators to set consumption tiers and cap usage down to the individual user. Furthermore, the platform maintains its model-agnostic nature, supports on-premises and air-gapped deployments, and integrates with major enterprise tools like Slack, GitHub, and SharePoint.

Key Points

  • The new North 2 platform offers granular token spending controls, giving enterprises precise budgetary oversight over agent usage.
  • It features an advanced orchestration system that manages multi-step tasks and allows for the sharing of proven agent workflows across an entire company.
  • North 2 enhances security with built-in guardrails to screen for PII and prevent prompt injection, while remaining model-agnostic for flexibility.

Why It Matters

This is a significant platform maturation announcement, moving AI agent deployment from experimental proof-of-concept to enterprise-grade operational reality. The focus on cost control, governance (guardrails), and integration depth signals a shift from 'can we build it?' to 'how do we run it reliably and affordably at scale?' This addresses the primary bottlenecks preventing widespread, mission-critical adoption of autonomous AI agents in regulated industries.

You might also be interested in