Advancing Sovereign Canadian AI Agents
TORONTO, June 26, 2025 (GLOBE NEWSWIRE) — kama.ai, Canada’s leader in Responsible Conversational AI, is pleased to announce that Cohere, the leading data security-focused enterprise AI company, is now available on kama.ai’s hybrid AI agent platform to power generative intelligence within GenAI’s Sober Second Mind® Hybrid AI Agents.
This integration deepens kama.ai’s commitment to provide trusted, human-governed knowledge delivery. This while harnessing high-quality generative outputs through Cohere’s secure enterprise LLMs and embedding models. Notably, both kama.ai and Cohere represent Canada’s growing Sovereign AI capabilities. This is an emerging important topic with enterprise clients.
Sovereign AI implies artificial intelligence technologies which are designed, developed, deployed, and governed within a specific country’s legal and ethical frameworks. This has become an issue concerning national control over data, algorithms, infrastructure, and compliance.
“Cohere strengthens our Hybrid AI Agent solution by providing a Canadian-built GenAI capability alongside our trusted knowledge graph AI technology,” said Brian Ritchie, CEO and Founder of kama.ai. “We see this as a very positive step forward. Combining these two technologies provides the most responsible and brand-safe AI systems in the industry.”
Trusted Collections for Enterprise Assurance
This new development allows kama.ai to connect and use Cohere’s Command family LLMs to deliver real-time creative responses. This while Cohere’s embedding and rerank models power the company’s Trusted Collections for enterprise Retrieval Augmented Generation (RAG). Trusted Collections ensure that only carefully chosen documents are selected and vectorized to inform generative drafts or real-time responses. This minimizes challenges with hallucinations, and other contamination from less relevant information. Draft Assist with kama’s Trusted Collections allows generative support while human experts validate critical information for the organization’s knowledge graph. This allows kama.ai systems to provide 100% accuracy and reliable adherence to brand sanctioned information.
This process means enterprise knowledge managers and administrators can maintain full control over what the AI Agent says publicly. It blends the speed and creativity of Cohere’s generative AI with the deterministic accuracy and factual safety of kama.ai’s graph-based enterprise knowledge.
A Stronger Canadian AI Ecosystem
CAISIC, Canada’s Artificial Intelligence Sovereignty and Innovation Cluster, praised this milestone as an advancement for Canada’s AI sector: “kama.ai’s integration with Cohere demonstrates that Canada’s AI innovators are capable of delivering collaborative solutions built with local expertise,” said Niraj Bhargava, Co-Founder and Co-Chair of CAISIC. “This direction aligns with the growing need for sovereign, ethical, and trusted AI technologies in the AI stack.”
Story Continues