One enterprise AI platform, Scotia Intelligence, rolled out to more than 71,000 employees and built to be model-agnostic, with a board technology committee, a public data-ethics function and a shareholder vote that shows investors want more disclosure.
How is Scotiabank using AI?
Scotiabank, which describes itself as one of the largest banks in North America by assets and reported C$1,548 billion of total assets at July 31, 2026, runs its AI programme through Scotia Intelligence, a unified data-and-AI platform launched on April 13, 2026. By June 10, 2026 more than 71,000 employees had access to its assistive AI tools and 5,500 engineers were using AI for coding; by July 31 staff had generated 14 million actions on it since the March 2026 launch, and at a September 2026 investor conference management said 95% of employees had adopted Scotia Navigator. The bank reports production results in client service (AI handles more than 40% of contact-centre queries) and commercial banking (about 90% of commercial emails are routed by AI, cutting manual work by 70%). CEO Scott Thomson told investors in May 2026 that Scotiabank is model-agnostic because no single model or vendor will dominate for long, and in July the bank became a founding member of the AI Consortium with Lightworks, Sun Life and TELUS to share an Agentic Control Plane. Governance is visible: a board technology committee, a data-ethics team, and a shareholder proposal asking for an AI report that won 22.21% of votes cast in April 2026.
Lead bank
The Bank of Nova Scotia
Total assets (group, as reported, Jul 31, 2026)
C$1,548bn total assets (reported in Canadian dollars)[8]
Charter
Canadian chartered bank; supervised by the Office of the Superintendent of Financial Institutions (OSFI), with US operations under US banking supervisors
Headquarters
Toronto, Canada
Flagship platform
Scotia Intelligence and Scotia Navigator
AI leadership
Scott Thomson · Tim Clark · Phil Thomas · Sandeep Kumar · Jenny Poulos · Shannon McGinnis · Dan Yungblut
- Platform: Scotia Intelligence (April 13, 2026) with Scotia Navigator for employees; new Notebooks, Create and Pages tools (June 10) and knowledge agents (July 31). More than 71,000 employees have access and the bank counts 14 million actions since March 2026.
- Production metrics disclosed: AI handles more than 40% of contact-centre client queries; about 90% of commercial emails are processed by AI, cutting manual work by 70%; 5,500 engineers use AI for coding (June 2026).
- Model-agnostic by design: on its Q2 2026 call management said no single model or vendor will dominate over an extended period, so models are selected on performance, security and cost; the sources reviewed do not name the models behind Navigator.
- Shared governance infrastructure: Scotiabank is a founding member of the AI Consortium (July 7, 2026) with Lightworks, Sun Life and TELUS, whose Agentic Control Plane processes more than two trillion tokens a month across members; Scotiabank expects to use it in its own AI governance.
- Oversight under pressure: a shareholder proposal for a report on AI in decision-making, credit underwriting and risk assessment received 22.21% support at the April 14, 2026 meeting, up from 12.19% a year earlier; the board recommended against it, citing existing governance.
- Spending: technology spend was about C$1.3 billion in Q1 2026 and C$1.5 billion in Q3 (up 16%); the CEO says AI is an important and growing part of it, but the bank does not break AI out separately.
- Cyber: the Q3 2026 report names frontier AI as a cyber threat and says the bank is moving to a real-time, automated security model with AI-enabled detection and response.
What is Scotia Intelligence and Scotia Navigator?
Scotia Intelligence, announced on April 13, 2026, is the bank's unified approach to data and AI: the capabilities, platforms and governance needed to deliver AI to employees and clients. Its employee-facing layer, Scotia Navigator, provides assistive AI inside the workplace, lets teams build and deploy custom assistants and research or analytics agents in a governed environment, and gives engineers AI coding assistance. Two months later the bank added Notebooks, Create and Pages, and on July 31 it added knowledge agents (Delivery Navigator, a Travel and Expense agent and a Procurement agent). On the May 2026 call CEO Scott Thomson said the bank chose a model-agnostic approach from day one, selecting models on performance, security and cost, and built on an enterprise data platform with agent deployment and continuous monitoring of models in production.[1][2][3][14][16]
What has Scotiabank done on AI, and when?
- Sep 9, 2026
- Aug 25, 2026
Q3 2026: expanded platform, higher tech spend, frontier-AI cyber risk[15][8]
Management cites the new Scotia Intelligence features, knowledge agents and the AI Consortium; technology spend rose 16% to C$1.5 billion; the quarterly report names frontier AI, including Anthropic's Claude Mythos Preview, as raising cyber-vulnerability exposure and describes a move to AI-enabled security.
- Jul 31, 2026
Knowledge agents added; 14 million actions[3]
Delivery Navigator, a Travel and Expense agent and a Procurement agent join Scotia Intelligence; more than 80% of the global workforce has completed at least one internal AI course.
- Jul 7, 2026
- Jun 10, 2026
Notebooks, Create and Pages; adoption figures[2]
More than 71,000 employees enabled for assistive AI tools, 5,500 engineers using AI for coding, a 30% quarter-over-quarter increase in employees using AI to reply to client questions, and mandatory AI risk training.
- May 27, 2026
Q2 2026 call: four principles of the AI approach[14]
Thomson sets out security, flexibility (model-agnostic, no single vendor), data and platform-first thinking, adding that the platform enables AI agents and continuous monitoring of models in production.
- Apr 14, 2026
AGM: AI oversight proposal wins 22.21%[10][9][11]
Shareholder Proposal No. 6, asking for a report on AI in senior-management decisions, risk assessment and credit underwriting, is opposed by the board but receives 22.21% of votes cast; the CEO's address presents Scotia Intelligence as the enterprise approach to data and AI.
- Apr 13, 2026
Scotia Intelligence and Scotia Navigator launched[1][16][17]
A unified enterprise approach to AI for employees and clients: assistive AI, custom assistants and agents, coding assistance, with governance built in. The bank reports AI handling over 40% of contact-centre queries and about 90% of commercial emails.
- Feb 24, 2026
Q1 2026 call: Ask AI volumes and a Tangerine AML pilot[12][13]
Scott Thomson says AI is an important and growing part of technology spending, Ask AI processed over 450,000 queries in the quarter, and a Tangerine AML pilot with an external partner cut existing alert volumes by 37%; the bank will build its own solution.
- Dec 2, 2025
2025 annual report: AskAI, Scotia chat and an AI risk factor[7]
The report says AskAI, the internal chatbot, had assisted more than 15,000 employees with over 750,000 questions, Scotia chat handles over 132,000 inquiries a month and AIDox automates commercial and small-business email processing. Its risk section adds a Rapid AI Adoption factor and an AI Risk Management Program.
- Nov 18, 2025
Phil Thomas takes the strategy and operating role with AI innovation in scope[21]
The Globe and Mail reports that Thomas, previously chief risk officer, becomes Group Head and Chief Strategy and Operating Officer, focused on client experience and AI innovation and overseeing data and analytics; Shannon McGinnis becomes chief risk officer.
Where does Scotiabank use AI today?
| System | What it does | Status | Use case |
|---|---|---|---|
| Scotia Navigator assistive AI platform[1][2] | Employee AI assistant for research, analytics and document work, with custom assistants and agents built in a governed environment; Notebooks, Create and Pages added in June 2026. | In production | Generative & agentic AI |
| Knowledge agents[3] | Delivery Navigator, Travel and Expense and Procurement agents answer employee questions from approved internal sources; the bank describes a shift from general-purpose assistance to agents and AI-enabled workflows. | Rolling out | Generative & agentic AI |
| Contact-centre AI and Ask AI[1][12][7] | AI handles more than 40% of client queries in contact centres; Ask AI gives branch and contact-centre staff policy and product guidance (450,000-plus queries in Q1 2026); Scotia chat handles over 132,000 inquiries a month. | In production | Customer-facing chatbots |
| Commercial email automation (AIDox)[1][7] | AI processes about 90% of commercial emails, routing them to the right teams and cutting manual work by 70%. | In production | Generative & agentic AI |
| Tangerine AML AI pilot[12] | An externally supported pilot cut existing alert volumes by 37%; the bank is designing its own solution to improve precision, reduce false positives and avoid vendor dependency. | Pilot | AML / KYC |
| AI coding assistance[1][2][20] | Scotia Navigator provides AI-powered coding assistance; 5,500 engineers used it by June 2026 and about 5,000 coders were described as active users in September. | In production | AI-generated code & coding agents |
| InvestIQ research tool[6] | Scotia Global Asset Management's tool summarizes quarterly earnings, finds patterns in disclosures and prepares for management meetings; time savings of up to 90% on some tasks. | In production | Generative & agentic AI |
| Data ethics and AI governance[1][9] | A dedicated Data Ethics team, a public Data Ethics Statement, review of AI use cases for fairness, transparency and accountability before launch, mandatory training and annual attestations; a board technology committee oversees AI and data-ethics frameworks. | In production | AI governance (general) |
| AI Risk Management Policy and model risk framework[9][7] | AI risk is anchored in the Model Risk Management framework and an AI Risk Management Policy; the bank says it is preparing for OSFI Guideline E-23, effective May 1, 2027. | In production | Model risk management |
| AI Consortium and Agentic Control Plane[4][3][18] | Shared control layer across models, agents, users and inference pipelines; Scotiabank is a founding member and expects to use the consortium's tools in its AI governance. | Announced | Third-party & vendor AI |
| AI-enabled security[8][14] | In response to frontier-AI risks the bank is moving to a real-time, automated security model with AI-enabled detection, automated response and scalable remediation, and uses AI-driven scanning and monitoring. | Rolling out | Cybersecurity |
Use-case links open the tracker's rules-by-use-case matrix: which documents govern each system type.
What has Scotiabank disclosed in numbers?
| Metric | Value | As of |
|---|---|---|
| Employees with access to Scotia Intelligence AI tools | More than 71,000[2] | Jun 10, 2026 |
| Engineers using AI for coding | 5,500[2] | Jun 10, 2026 |
| Actions generated on Scotia Intelligence since March 2026 launch | 14 million[3] | Jul 31, 2026 |
| Employees who have adopted Scotia Navigator (management statement) | 95%[20] | Sep 9, 2026 |
| Workforce completing at least one internal AI course | More than 80% of global workforce; over 90% of directors and above[3] | Jul 31, 2026 |
| Contact-centre client queries handled by AI | More than 40%[1] | Apr 13, 2026 |
| Commercial emails processed by AI | About 90%, cutting manual work by 70%[1] | Apr 13, 2026 |
| Ask AI queries in the quarter | More than 450,000[12] | Jan 31, 2026 |
| Tangerine AML pilot: reduction in existing alert volumes | 37%[12] | Feb 24, 2026 |
| Technology spend, Q3 2026 | C$1.5 billion, up 16% year over year[15] | Jul 31, 2026 |
| Shareholder support for AI-oversight report proposal | 22.21% of votes cast (12.19% the year before)[10][9] | Apr 14, 2026 |
| Workdays saved by AI tools in wealth management (management statement) | 24,000 over about a quarter and a half[20] | Sep 9, 2026 |
| Total assets | C$1,548bn (Canadian dollars)[8] | Jul 31, 2026 |
“Our approach to AI is designed to take us from isolated AI use cases to AI that is embedded across our processes, decision-making, and client interactions in a trusted, efficient, and effective manner.”[14]
Scott Thomson · President and Chief Executive Officer · May 27, 2026
“The AI landscape is evolving rapidly, we believe that no single model or vendor will dominate over an extended period of time.”[14]
Scott Thomson · President and Chief Executive Officer · May 27, 2026
“AI is an important and growing part of our total technology spending.”[12]
Scott Thomson · President and Chief Executive Officer · Feb 24, 2026
“The next phase of AI is about moving from information access to intelligent action”[3]
Tim Clark · Group Head and Chief Information Officer · Jul 31, 2026
“If you look at AI, frankly, and how it has evolved, everybody has moved from pilot through to some level of agent-based workflows”[18]
Tim Clark · Group Head and Chief Information Officer · Jul 7, 2026
Who runs AI at Scotiabank?
Which regulators govern Scotiabank's AI?
| Authority | Why it matters here | Documents |
|---|---|---|
| OSFI | Scotiabank's lead federal supervisor; the bank's 2026 proxy circular says it is subject to OSFI requirements on model, cyber, third-party and outsourcing risk and is preparing for Guideline E-23 (model risk management), which takes effect on May 1, 2027 and covers AI models. | |
| Financial Consumer Agency of Canada | The bank's proxy circular names FCAC market-conduct obligations among the AI-related requirements it is subject to, relevant to AI used in client service and sales. | |
| Federal Reserve | Scotiabank's US corporate and capital-markets business, which management says it is expanding, is supervised by US banking agencies including the Federal Reserve; third-party and model-risk guidance frames group AI platforms used by US staff. | SR 23-4 SR 26-2 |
| Basel Committee | An enterprise data platform built to be 'ready for AI consumption' and shared control infrastructure with other firms engage BCBS 239 risk-data aggregation principles and the Basel third-party risk principles. | BCBS 239 BCBS Third-Party Risk Principles (Dec 2025) |
What does the public record suggest about Scotiabank's AI strategy?
Analysis by BankingNewsAI from the sources cited on this page. Observations, not advice.
01
Governance is a disclosed workstream, and investors are asking for more
Scotiabank describes a board technology committee, an AI & Ethics Governance team and an Ethics Assistant review of every AI use case, yet a proposal for a standalone AI report drew 22.21% support, almost double the prior year's 12.19%. The gap suggests investors want outcome-level disclosure that the bank's current statements do not give.
02
Model-agnostic plus a shared control plane lowers lock-in
Choosing models by performance, security and cost, and co-owning an Agentic Control Plane with Sun Life, TELUS and Lightworks, points to a strategy of controlling the layer around the models rather than depending on one vendor. It also concentrates governance work on monitoring and agent permissions, which supervisors increasingly examine.
03
Adoption figures measure different things
The bank reports 71,000 employees 'enabled' (June), 14 million actions (July) and 95% adoption (September, a management statement at an investor conference). Access, activity and adoption are different measures, and the public record has no disclosed AI spending or net benefit figure, only production metrics such as 40% of contact-centre queries and 70% less manual email work.
04
A bank naming frontier AI in its cyber risk
Few quarterly reports name a specific frontier model when describing cyber exposure. Scotiabank's Q3 2026 report does, and pairs it with a move to AI-enabled detection and response, tying its AI platform investment to defence as well as productivity.
What is Scotia Intelligence?
Scotia Intelligence is Scotiabank's unified approach to data and AI, announced on April 13, 2026. It combines the capabilities, platforms and governance needed to deliver AI to employees and clients. Its employee-facing component, Scotia Navigator, provides assistive AI, custom assistants and agents, and coding assistance. By June 2026 more than 71,000 employees had access to the tools.
Which AI models does Scotiabank use?
Scotiabank says it is model-agnostic: CEO Scott Thomson said in May 2026 the bank selects models based on performance, security and cost because no single model or vendor will dominate for long. The sources reviewed for this page do not name the models behind Scotia Navigator.
How does Scotiabank govern its AI?
AI risk sits within the Model Risk Management framework and an AI Risk Management Policy, overseen by a board technology committee and the risk committee, with a Data Ethics team that reviews AI use cases for fairness, transparency and accountability before launch. Employees complete mandatory AI risk training and annual attestations, and the bank is a founding member of the AI Consortium building shared control tools.
Has Scotiabank disclosed results from AI?
Yes, as operating metrics rather than a dollar benefit: AI handles more than 40% of contact-centre queries, about 90% of commercial emails are routed by AI with 70% less manual work, a Tangerine AML pilot cut alert volumes by 37%, and management said AI tools saved 24,000 workdays in wealth management over about a quarter and a half. It has not published a total AI investment figure.
- Scotiabank Launches Scotia Intelligence, Empowering Employees and Accelerating Enterprise AI Adoption
- Scotiabank Accelerates Enterprise Adoption of AI with New Scotia Intelligence Capabilities
- Scotiabank Equips Employees with AI-Powered Knowledge Agents Through Scotia Intelligence
- Lightworks, Scotiabank, Sun Life and TELUS launch AI Consortium to jointly build critical AI control infrastructure in Canada
- VIDEO: AI Unpacked – Scotia Intelligence accelerates the Bank’s AI evolution and adoption
- VIDEO: AI Unpacked – How InvestIQ, Scotia’s AI tool, is enhancing investment research
- Scotiabank 2025 Annual Report - English
- Scotiabank Q3 2026 Report to Shareholders-English
- Annual and Special Meeting of Shareholders: Management Proxy Circular, April 14, 2026
- Report of Voting Results, Annual and Special Meeting of Shareholders held April 14, 2026
- Scotiabank's 194th Annual and Special Meeting of Shareholders: CEO address
- Bank Nova Scotia Halifax Pfd 3 (BNS) Q1 FY2026 earnings call transcript
- Bank of Nova Scotia Q1 Earnings Call Highlights
- Bank Nova Scotia Halifax Pfd 3 (BNS) Q2 FY2026 earnings call transcript
- Scotiabank (BNS) Q3 2026 Earnings Call Transcript
- Scotiabank Equips Global Staff With New Scotia Intelligence AI
- Canada’s Scotiabank preps for its AI future
- Scotiabank teams up with other Canadian companies to meet their AI needs faster, cheaper
- Banks say AI is allowing them recover unpaid loans, cut call centre volumes and save workdays
- Scotiabank at annual Financials Summit: early ROE goal met
- Scotiabank names new chief risk officer amid management shuffle
When Scotiabank moves on AI, you'll read it here first.
when one of the banks we cover moves on AI, it is in the next morning's brief · six sourced stories · 7 am ET · free
plus every tracker, bank and agent page update, the morning after · leave any morning