← All use cases

Bell Cyber (Bell Canada) · Toronto, ON

Bell Cyber and Cohere put a Canadian-hosted AI model into its security operations centre

Security & FraudLegal/ComplianceProductivitySovereign AI

Live in production (Sept 2026)
Canadian-hosted processing
No quantified results published

What was done

Bell Cyber and Cohere built a cybersecurity-specific language model trained on Bell Cyber's own data and know-how, and are integrating it into Bell Cyber's Autonomous Security Operations Centre. It summarizes complex security information, pulls out the context that matters for an investigation and supports standardized investigative steps, while analysts keep the final say. It runs on Bell AI Fabric so processing and sensitive data stay in Canada.

Outcome

Bell's September 2026 announcement says the model is running in production and is the first customer-facing product of the Bell–Cohere partnership; it describes reduced time spent gathering and interpreting information but publishes no figures.

Evidence: Company press release (Bell Canada). Figures are as reported by the source and have not been independently audited by AgentsNorth.

Prompt Reconstructed by AgentsNorth

The source didn't publish its prompt. This is our reconstruction of a starting prompt for a similar workflow, not the company's actual prompt.

You are an investigation assistant in a security operations centre. Given this alert and the related endpoint, identity, network and cloud logs, summarize what happened in time order, list the affected assets and accounts, highlight the evidence that points to malicious activity and anything that suggests a false positive, and suggest the next investigative steps. Cite the log entries you used. Do not take containment actions; an analyst decides.

Related use cases

Get new Canadian AI agent use cases each week

New use cases, one prompt to steal and one failure to avoid. Free. Unsubscribe anytime.