Recap: Solytics Partners AI & MRM Conclave Toronto 2026: Advancing Enterprise AI Governance and Agentic AI

Advancing Enterprise AI Governance and Model Risk Management in the Age of Agentic AI

Solytics Partners successfully hosted the AI & MRM Conclave - Toronto 2026, bringing together senior leaders from banks, insurers, and financial institutions across Canada. The conclave focused on how organizations can evolve Model Risk Management (MRM) and AI Governance frameworks to responsibly govern traditional models, AI/ML, GenAI, and emerging Agentic AI systems under a unified enterprise-wide governance strategy.

Against the backdrop of evolving regulatory expectations including OSFI E-23 and global AI governance frameworks, discussions centered on building scalable governance, strengthening validation practices, and operationalizing responsible AI adoption across financial institutions.

Regulatory & Governance Landscape: From Model Governance to AI System Governance

The opening session explored how regulatory expectations are shifting from traditional model governance to enterprise-wide AI system governance. As financial institutions expand AI adoption across GenAI and Agentic AI use cases, governance frameworks must evolve to manage end-to-end AI lifecycle risks, including data, models, infrastructure, and decision-making agents.

The discussion emphasized alignment with global regulatory expectations and highlighted the growing need for continuous monitoring, lifecycle governance, and automation-driven oversight to ensure audit-ready AI deployment.

Workshop: Validation & Testing of AI/ML, GenAI, and Agentic Systems by Agus Sudjinatio, Senior Advisor, Solytics Partners

The interactive workshop focused on strengthening validation and testing frameworks for modern AI systems. The session covered explainability, bias detection, robustness testing, and continuous monitoring - critical components for governing AI in regulated environments.

Participants explored how validation practices must evolve for probabilistic AI and multi-agent workflows, including structured testing approaches, challenger models, performance benchmarking, and governance-driven validation pipelines.

Spotlight Session: Verification Layer for LLMs and Agentic AI

A key highlight of the conclave was the introduction of a Verification Layer for LLM-powered Agentic AI, designed to improve determinism, reliability, and governance of agent-driven systems.

The verification layer introduces structured validation checkpoints across agent workflows, enabling reasoning traceability, output validation, policy enforcement, and human-in-the-loop governance. This approach transforms Agentic AI from probabilistic automation into deterministic, regulator-ready decision-support systems.

The session demonstrated how verification-driven controls help organizations enforce governance policies, validate outputs, and maintain continuous oversight across AI workflows.

Technology Showcase: Agent-Driven Governance Architecture

Solytics Partners showcased its integrated AI governance ecosystem powered by Vault (Governance Layer) and NIMBUS (Execution Layer), connected through an intelligent Agent Layer. This architecture enables continuous governance across the AI lifecycle, from model onboarding to validation, monitoring, and regulatory reporting.

Specialized governance agents demonstrated included governance and compliance agents, validation and benchmarking agents, monitoring and alerting agents, workflow and approval agents, data quality agents, and explainability agents. Together, these agents enable automated governance, lifecycle monitoring, and regulator-ready assurance.

Panel Discussion: MRM Best Practices in the Age of AI

The conclave concluded with a high-level panel discussion featuring senior industry leaders who shared practical insights on evolving MRM and AI governance frameworks. Discussions focused on enterprise-wide governance, lifecycle automation, risk management for GenAI and Agentic AI, and building regulator-ready AI governance capabilities.

Panelists emphasized that organizations must transition from siloed model governance to enterprise-wide AI governance supported by automation, intelligent orchestration, and verification-driven controls.

Key Takeaways

The Toronto Conclave reinforced a clear industry direction, as AI adoption accelerates, financial institutions must evolve toward enterprise-wide governance frameworks that combine model risk management, agentic orchestration, and verification-driven controls.

With regulatory scrutiny increasing and AI capabilities expanding, verification layers, continuous governance, and automated lifecycle oversight will play a critical role in enabling responsible and scalable AI adoption across financial institutions.