KPMG Canada announced the launch of its Agentic AI Engine on May 1, 2025, marking one of the country’s most significant corporate investments in autonomous artificial intelligence technology. The system enables organizations to deploy digital agents that operate independently, make decisions, and execute tasks without constant human intervention.

These AI agents differ from conventional automation tools. They analyze data, determine optimal courses of action, and carry out complex workflows across multiple systems while adapting to changing conditions in real time.

KPMG in Canada Unveils Agentic AI Engine to Drive Business Efficiency

The Agentic AI Engine represents KPMG’s platform for developing, testing, and deploying autonomous AI solutions specifically tailored for Canadian businesses. The system already powers several specialized agents across different sectors.

Stephanie Terrill, Canadian Managing Partner of Digital and Transformation at KPMG Canada, emphasized the technology’s significant potential. She stated that agentic AI represents the most powerful form of artificial intelligence developed to date.

The firm positioned this launch against Canada’s economic backdrop of productivity challenges and trade uncertainties, framing autonomous AI as a strategic response to competitive pressures.

Walter Pela, AI Client and Market Development Lead for KPMG Canada, cited internal research showing nearly 90 percent of Canadian business leaders view agentic AI as essential for maintaining competitiveness and reducing operational costs.

The Engine facilitates partnerships with industry leaders, AI research institutions, and technology providers to accelerate knowledge sharing and practical implementation of autonomous systems.

Projected Productivity Boost of 30% for Canadian Businesses with Agentic AI

KPMG forecasts that Canadian organizations implementing agentic AI could achieve efficiency gains up to 30 percent within three years. This projection assumes teams and individuals work alongside digital agents rather than being replaced by them.

The productivity calculation accounts for reduced time spent on administrative tasks, faster decision cycles, and improved accuracy in complex processes that currently require extensive manual review.

These gains extend beyond simple task automation. Agentic systems can coordinate across departments, prioritize work based on real-time conditions, and execute multi-step processes that previously required human orchestration at each stage.

The technology’s impact varies by sector and use case. Financial services, professional services, and organizations with high document processing volumes show the strongest potential for immediate gains.

However, realizing these benefits requires organizations to redesign workflows around agent capabilities, retrain staff to work effectively with autonomous systems, and establish governance frameworks for AI oversight and accountability.

Real-World Applications of Agentic AI in Canadian Organizations

KPMG has deployed multiple specialized agents through its Agentic AI Engine, each designed to handle specific business functions autonomously.

An underwriting agent processes insurance claims and loan applications by automatically reviewing documents, conducting compliance checks, and flagging exceptions. KPMG reports this agent reduces administrative task time by approximately 40 percent.

The customer service agent handles inquiry triage, personalizes product offers, and processes returns without human intervention for routine cases. This frees thousands of workforce hours for complex customer issues requiring judgment and relationship management.

A risk management agent continuously monitors organizational data streams and external sources to identify emerging risks. It generates risk assessments autonomously and provides real-time alerts when threat levels exceed defined thresholds.

KPMG also developed an assessment agent that analyzes meeting transcripts, data models, and survey responses to identify improvement opportunities within organizations. This agent synthesizes information across multiple formats and sources to surface actionable insights.

For private equity clients, an ESG due diligence agent evaluates environmental, social, and governance risks during investment screening. The agent reviews regulatory filings, news sources, and corporate disclosures to flag material ESG concerns before deals close.

The firm embedded agents into KPMG Clara, its global audit platform, to deliver deeper insights and standardize workflows. Sebastian Distefano, Canadian Managing Partner for Audit and Assurance, noted these agents help audit professionals respond more effectively to risks while improving consistency and quality.

In tax operations, KPMG launched Ignition Tax to deploy AI agents through One Port, the firm’s centralized cloud hub for tax compliance and advisory work. These agents operate under human verification systems to ensure professional oversight.

Lucy Iacovelli, Canadian Managing Partner for Tax and Legal, explained that agents enhance professional workflows but maintain human accountability through built-in review protocols.

Scotiabank implemented a similar approach in its Commercial Banking division using AIDox, an AI document analysis system developed in-house and patented. The bank trained the system on complex, irregular client requests ranging from credit line withdrawals to transaction investigations.

AIDox now handles roughly 90 percent of the 1,500 daily emails to Commercial Banking’s centralized address. The system reads each request, determines the required action, routes it to the correct department, and creates a case for processing.

Lawrence Engel, Scotiabank’s Vice President of Business Banking Operations, reported that AIDox reduced routing time from one to two hours down to minutes while improving accuracy.

This deployment allowed Scotiabank to redeploy 70 percent of the team previously responsible for email routing into higher-value roles. Client satisfaction scores increased consistently over the deployment period, though Engel noted multiple factors contributed to this improvement.

These implementations demonstrate agentic AI’s capacity to handle variable, complex workflows that previously resisted automation due to their unpredictability and need for contextual judgment.

KPMG’s Strategic Collaboration to Accelerate AI Adoption in Canada

KPMG structured its Agentic AI Engine to facilitate collaboration across Canada’s AI ecosystem rather than operating as an isolated development effort.

The Engine connects industry leaders, academic research institutes, and technology alliance partners to share knowledge on agentic AI advancements. This approach mirrors Canada’s broader strategy of leveraging its research strengths in institutions like Mila, the Vector Institute, and Alberta Machine Intelligence Institute.

KPMG Canada partnered with Microsoft Canada to establish the Skills Development Centre, offering training modules on agentic AI for executives and business leaders. The program provides both complimentary and custom-tailored courses on AI opportunities and challenges.

More than 53,000 Canadians received professional training in AI and cybersecurity through the Skills Development Centre as of May 2025. This educational initiative addresses the skills gap as organizations move from generative AI adoption to autonomous system deployment.

The firm also enhanced KPMG Kleo, its proprietary generative AI platform, with agentic capabilities. This allows KPMG professionals to build their own task-specific agents, testing concepts internally before deploying solutions for clients.

Terrill described this approach as treating KPMG as ‘Client Zero,’ ensuring solutions are grounded in practical experience before external deployment. This methodology addresses reliability concerns around agentic systems, which research indicates still achieve professional-level task completion only a fraction of the time.

Canada’s federal government reinforced this momentum in June 2026 with its AI for All strategy, committing billions toward AI adoption and domestic capability building. The plan targets 200 billion dollars in GDP growth and 250,000 new jobs over five years.

Policymakers recognized that agentic AI introduces distinct governance challenges. Because these systems act autonomously rather than simply providing recommendations, frameworks must address accountability for actions taken, auditability of decision logic, and potential unintended interactions between multiple agents operating simultaneously.

This focus on governance aligns with emerging regulatory frameworks that pair innovation incentives with safety-first principles grounded in transparency and public trust.

Nearly 16.5 million dollars in May 2026 went to AI-focused businesses in the Greater Toronto Area to accelerate commercialization and adoption, demonstrating government support for translating research into deployed systems.

Broader Implications for Canada’s Technology Sector

The shift from generative to agentic AI represents more than incremental improvement. It changes the fundamental relationship between humans and AI systems from question-and-answer interactions to delegation of entire workflows.

This transition raises the stakes for reliability, safety, and trust. When AI systems make recommendations, humans retain final decision authority. When they act autonomously, organizations must establish guardrails, monitoring systems, and intervention protocols before problems cascade.

Canada’s approach emphasizes building these safeguards into deployment from the start rather than addressing them reactively. KPMG’s human verification systems and Scotiabank’s anomaly review processes demonstrate this principle in practice.

The technology also reshapes workforce dynamics. Rather than replacing workers, early implementations show agents handling routine complexity while redirecting human effort toward judgment-intensive tasks, relationship management, and strategic work.

However, this transition requires significant organizational change management. Workers must learn to oversee agent performance, intervene when systems exceed their capabilities, and contribute to continuous improvement of agent logic based on real-world outcomes.

The economic implications extend beyond individual organizations. If Canada’s productivity gains materialize at projected levels, agentic AI could meaningfully address the country’s long-standing productivity gap relative to peer economies.

Yet success depends on broad adoption across sectors, not just early movers. Smaller organizations and sectors with less technology investment may struggle to implement these systems without support, potentially widening competitive gaps.

The security landscape also evolves with agentic deployment. Autonomous systems present new attack surfaces, and compromised agents could execute harmful actions at scale before detection. This makes cybersecurity architecture critical to safe implementation.

Canada’s integrated ecosystem of research institutions, commercial ventures, and government support positions it well for this transition. The country’s reputation for trustworthy AI development may become a competitive advantage as global organizations seek partners for sensitive autonomous system deployments.

Frequently Asked Questions

What is agentic AI and how does it differ from traditional AI?

Agentic AI systems operate independently to identify, plan, and execute tasks without requiring human prompts for each action. Unlike generative AI tools that respond to user queries with text or images, agentic AI pursues defined goals by sequencing actions, interacting with multiple software systems, and adapting to changing conditions. For example, while generative AI might draft an email when asked, an agentic system could identify the need for communication, compose the message, send it at optimal timing, and follow up based on recipient response without human intervention at each step.

How will KPMG’s investments in agentic AI impact the Canadian economy?

KPMG projects its Agentic AI Engine could help Canadian organizations achieve productivity gains up to 30 percent over three years by enabling teams to accomplish more with existing resources. If adopted broadly, these efficiency improvements could address Canada’s productivity challenges and contribute to economic growth during a period of trade uncertainty. The federal government’s AI for All strategy targets 200 billion dollars in GDP growth and 250,000 new jobs over five years through AI adoption, though realizing these benefits depends on successful implementation across sectors and organization sizes.

What types of organizations can benefit from KPMG’s Agentic AI Engine?

Organizations with high volumes of document processing, complex multi-step workflows, and regulatory compliance requirements show the strongest potential for agentic AI benefits. This includes financial services firms handling underwriting and risk assessment, professional services organizations managing client engagements, retailers processing customer inquiries and returns, and private equity firms conducting due diligence. However, successful implementation requires organizations to redesign processes around agent capabilities, establish governance frameworks, and train staff to work effectively alongside autonomous systems rather than simply deploying technology into existing workflows.

Conclusion

KPMG Canada’s launch of its Agentic AI Engine in May 2025 represents a pivotal shift in how Canadian organizations approach artificial intelligence. Rather than viewing AI as a tool that responds to prompts, agentic systems act as autonomous collaborators that execute complex workflows independently.

The technology’s potential to deliver 30 percent productivity gains addresses urgent economic challenges for Canadian businesses. Early implementations across financial services, professional services, and other sectors demonstrate that these improvements are achievable when organizations commit to process redesign and workforce adaptation.

However, realizing this potential requires more than technology deployment. Organizations must build governance frameworks that ensure accountability, establish monitoring systems that detect when agents exceed their capabilities, and invest in training that enables workers to supervise and improve autonomous systems effectively.

Canada’s integrated approach combining corporate investment, academic research, and government support creates conditions for leadership in trustworthy agentic AI development. As autonomous systems become more prevalent globally, the country’s emphasis on safety, transparency, and human oversight may prove as valuable as its technical capabilities.