A group-level AI programme run from Toronto around Layer 6, its in-house AI lab, with a C$1 billion annual AI value target (half revenue, half cost), a first agentic AI model in mortgages, a C$25 million Cohere collaboration and an Evident AI Index rank that slipped from ninth to 13th in 2025 and recovered to 10th in 2026.
How is TD Bank using AI?
The Toronto-Dominion Bank, which reported C$2,111.9 billion of total assets at 31 July 2026, runs its AI programme from the group centre in Toronto and measures it in dollars: CEO Raymond Chun set a target of C$1 billion a year of value from AI over the medium term, half revenue and half cost, at the September 2025 Investor Day, and said on 27 August 2026 the bank had 'essentially hit' its fiscal 2026 target of C$200 million three quarters into the year, with C$195 million delivered by then. The build is centred on Layer 6, the AI lab TD bought in 2018, which produced the TD AI Prism foundation model (June 2025) and the bank's first agentic AI model (May 2026), which cut mortgage pre-adjudication from about 15 hours to under three minutes. TD's AI rank fell to 13th from ninth in the 2025 Evident AI Index, which also noted a 35% rise in AI-focused staff, and rose to 10th in the 2026 index published on 6 October 2026. On 22 September 2026 it announced up to C$25 million over three years for a Layer 6 and Cohere collaboration, and on 3 September 2026 it gave enterprise AI responsibility to Paul Whitehead. TD's US banking arm, under an asset cap after its 2024 anti-money-laundering settlement, is covered separately at /banks/td-bank-us.
Lead bank
The Toronto-Dominion Bank
Total assets (group, as reported, Jul 31, 2026)
C$2,111.9 billion total assets[1]
Charter
Canadian Schedule I bank; federally regulated financial institution supervised by the Office of the Superintendent of Financial Institutions (OSFI), with US banking subsidiaries supervised by the OCC and the Federal Reserve
Headquarters
Toronto, Canada
Flagship platform
Layer 6, TD AI Prism and the agentic RESL model
AI leadership
Raymond Chun · Luke Gee · Maksims Volkovs · Paul Whitehead · Vlad Shpilsky · Sumee Seetharaman · Ajai Bambawale · Rizwan Khalfan · Mohit Veoli · Taylan Turan · Licenia Rojas · Jesse Cresswell · Chad Koziel · Dan Bosman · Ted Paris · Kiran Vuppu · Eric Huber · Jo Jagadish · Sona Mehta
- Value target: C$1 billion a year from AI over the medium term, half from cost savings and half from revenue; about C$170 million of AI value was extracted in 2025, C$195 million in the first three quarters of fiscal 2026 against a C$200 million target, and Chun told investors on 9 September 2026 he saw 'sizeable upside' to the C$1 billion.
- Layer 6 (acquired 2018) is the engine: TD AI Prism (June 2025), a causal foundation model, the agentic mortgage model (May 2026) and, from September 2026, a funded collaboration with Cohere at the MaRS Discovery District.
- First wave of scaled AI, as named on the Q3 2026 call: retail end-to-end credit, the software development life cycle and contact centres; TD Insurance leads Canada in AI vehicle-damage estimation and targets C$150 million of claims-cost savings.
- Scale of use: more than 40,000 colleagues on Copilot, more than 7,000 engineers using AI (the most active with a 29% increase in throughput), a generative knowledge-management assistant in more than 1,000 Canadian branches and supporting 20,000-plus client-facing colleagues.
- Benchmark: Evident ranked TD 13th of 50 banks in October 2025, down from ninth, citing competition in innovation and talent; Evident said the dollar targets 'suggest that the bank could well regain ground'; on 6 October 2026 Evident placed TD 10th, citing gains in innovation, talent, leadership and transparency.
- Governance: Responsible AI Principles (seven commitments, 30 July 2026) sit inside an AI Risk framework overseen by an AI Risk Management team and a Trustworthy AI team that tests models for fairness and explainability before launch and monitors them after.
- Organisation: Paul Whitehead takes enterprise AI from 4 September 2026; Vlad Shpilsky leads Global Technology and Solutions; Luke Gee, Chief Analytics and AI Officer in TD's releases to July 2026, has left for Scotiabank (its Chief Data and AI Officer by 21 October 2026); COO Taylan Turan is leaving.
- US bank: TD Bank, N.A. and its AML remediation are covered at /banks/td-bank-us; this page covers the group's Canadian and enterprise AI record. The Canadian peer with a similar AI value target is covered at /banks/rbc.
- Restructuring and headcount: the programme announced in May 2025 (about 2,000 employees, 2% of the workforce) concluded on 31 January 2026 with C$886 million of pre-tax charges and about C$775 million of annual savings, including a workforce reduction of about 3%; the Q1 2026 filing lists severance, real estate and business wind-downs and does not attribute the charges to AI.
- Back in the top ten: Evident's 2026 AI Banking Index (6 October 2026) ranked TD 10th of 50, up three places; it noted C$195 million of AI value in nine months, a measurable impact reported for 20% to 25% of TD's use cases, the second-largest AI research team in the index and use of the agentic mortgage tool on 100% of new personal mortgages.
- Layer 6 is being scaled: the Globe and Mail reported in April 2026 that it was taking over a full floor at MaRS, growing from about 90 staff to about 240, while TD plans to raise the share of its technology budget spent on innovation and development to 45% from 34% over three years.
- Agentic work is widening: on 13 October 2026 TD announced agentic capabilities for mortgage funding and discharge, with colleagues finishing the initial funding review in up to half the time; on 9 September Chun said mortgage funding costs about C$187 a file and is heading to about C$47, and discharges about C$22 to C$23, heading to about C$11.
- Vendors named in the record: TD Securities' assistant runs on OpenAI's enterprise GPT through TD's Microsoft Azure relationship; GitHub Copilot was piloted with engineers from September 2023; Chun said in April 2026 TD does some work with Anthropic but was not in the limited Claude Mythos preview, and that TD plans a personal AI agent for each front-line colleague.
- AI risk ownership (2025 annual report): primary accountability sits with senior management of each business; Compliance, Data Risk, Model Risk, Technology and Cybersecurity Risk, Privacy and Third-Party Risk oversee it under a three-lines-of-defence model, and the report names generative and agentic AI as sources of regulatory, bias, operational, fraud, cyber, privacy and third-party risk.
What is Layer 6, TD AI Prism and the agentic RESL model?
Layer 6, the Toronto AI start-up TD acquired in 2018, is the bank's AI research and development centre; its Toronto and New York offices act as TD's AI centre of excellence. Its team of more than 130 built TD AI Prism, a predictive foundation model launched in June 2025 that replaces many single-purpose models and, according to Chief AI Scientist Maksims Volkovs, improved predictive accuracy by 20% to 30%, lead quality by 53% and campaign response rates by more than 40%. In May 2026 Layer 6 and the bank's technology, data, lending and risk teams launched TD's first agentic AI model, which generates application summary memos for mortgage and home equity underwriters and cut pre-adjudication from an average of 15 hours to under three minutes. Around these sit generative knowledge-management assistants for client-facing colleagues (20,000-plus by August 2026), Copilot for more than 40,000 colleagues and AI-assisted coding for more than 7,000 engineers.[6][13][9][3][4]
What AI partnerships and patents does TD Bank Group have?
TD builds its main models at Layer 6 and partners selectively. Named partners are OpenAI (the TD Securities assistant), Microsoft (Azure and GitHub Copilot), Cohere (up to C$25 million over three years) and the MIT Media Lab; Chun has said TD does some work with Anthropic without giving scope. Patents are assigned to The Toronto-Dominion Bank and include granted claims on transformer training, private-data model training and generative AI for software architecture, security testing, benefit letters and goal-based imagery. TD has not disclosed contract values for OpenAI, Microsoft or Anthropic.[32][31][33][12][35][29][44][49][46][47][50][48]
| Item | What it is | Status | Date |
|---|---|---|---|
| OpenAI | Enterprise GPT models behind the TD AI Virtual Assistant for TD Securities sales, trading and research, built with Layer 6 | In production | 2025-07-08[31][32] |
| Microsoft (Azure, GitHub Copilot) | Cloud partnership that made OpenAI enterprise models the natural platform; GitHub Copilot piloted with engineers from September 2023 | In production | 2025-10-09[32][33] |
| Cohere | Layer 6 and Cohere collaboration, up to C$25 million over three years, on knowledge management, security and responsible adoption | Announced | 2026-09-22[12] |
| Anthropic | Chun: TD does some work with Anthropic; TD was not in the limited Claude Mythos preview (scope and value not disclosed) | In use (scope not disclosed) | 2026-04-16[29] |
| MIT Media Lab sAIpien | Founding financial-services collaborator in a programme on trust, auditability and human-AI collaboration | Announced | 2025-10-28[35] |
| US 12,079,351 B2 | Applying trained AI processes to encrypted data to predict a credit score in real time | Granted | 2024-09-03[43] |
| US 12,321,861 B2 | Initialisation of parameters for machine-learned transformer architectures (Volkovs, lead inventor) | Granted | 2025-06-03[44] |
| US 12,326,856 B2 | Using generative artificial intelligence to improve user interactions | Granted | 2025-06-10[45] |
| US 12,399,687 B2 | Generating software architecture from conversation with a generative AI model | Granted | 2025-08-26[46] |
| US 12,517,812 B2 | Security testing based on generative artificial intelligence | Granted | 2026-01-06[47] |
| US 12,592,301 B2 | Prompt engineering and generative AI for goal-based imagery | Granted | 2026-03-31[48] |
| US 12,639,628 B2 | Distributed model training with collaboration weights for private data sets (Cresswell, lead inventor) | Granted | 2026-05-26[49] |
| US 12,657,566 B2 | Generative AI letter of confirmation of benefits | Granted | 2026-06-16[50] |
| US 2025/0077397 A1 | Large language model training for test case generation | Published application | 2025-03-06[51] |
| US 2025/0119396 A1 | Charge card knowledge chatbot using a large language model | Published application | 2025-04-10[52] |
What has TD Bank Group done on AI, and when?
- Oct 13, 2026
- Oct 6, 2026
Evident AI Banking Index 2026: TD returns to 10th[27]
Up from 13th; Evident cites improvement in innovation, talent, leadership and transparency, a measurable impact for 20% to 25% of use cases and the second-largest AI research team among the 50 banks.
- Sep 22, 2026
Layer 6 and Cohere: up to C$25 million over three years[12]
A strategic collaboration on knowledge management, security and responsible adoption, with a dedicated Cohere team working alongside Layer 6 at MaRS in Toronto.
- Sep 9, 2026
Chun: 'sizeable upside' to the C$1 billion target[15]
At the Scotiabank Financials Summit the CEO says an AI agent for collections lifted the customer connect rate to between 20% and 25% from 7% and that the mortgage agent is transferable to small business banking and auto finance.
- Sep 3, 2026
Enterprise AI moves under Paul Whitehead[11]
A senior-leadership reshuffle gives Whitehead responsibility for enterprise AI from 4 September, keeps Vlad Shpilsky at the head of Global Technology and Solutions and has COO Taylan Turan leave.
- Aug 27, 2026
- Jul 30, 2026
Responsible AI Principles introduced[10]
Seven enterprise-wide commitments embedded in the AI Risk framework, with model risk and privacy impact assessments and independent oversight reviews; the Trustworthy AI team monitors deployed models.
- Jul 16, 2026
OSFI bulletin on generative and agentic AI[20]
OSFI sets out sound practices for generative and agentic AI under its technology, operational-resilience and third-party guidelines (B-13, E-21, B-10), pointing to E-23 for model risk.
- May 28, 2026
Q2 2026 call: almost C$145 million of value; Copilot for 40,000[4]
Chun reports almost $145 million of AI value at the half-year, over 40,000 colleagues on Copilot, over 7,000 engineers using AI and, in insurance, Canada's first client-facing generative AI assistant from a home and auto insurer.
- May 21, 2026
First agentic AI model launched in mortgages and HELOCs[9]
Layer 6's model classifies documents, calculates income and checks policy before writing an underwriter memo, cutting pre-adjudication from an average of 15 hours to under three minutes; the Trustworthy AI team oversees it.
- May 1, 2026
TD Insurance launches a client-facing generative AI assistant[26]
The unauthenticated TDI Virtual Assistant, built with Layer 6 and Global Technology and Solutions and trained on website content, answers home, auto and small-business questions in English and French; TD calls it the first among Canada's Big Five.
- Apr 18, 2026
Layer 6 to grow from about 90 staff to about 240 at MaRS[28]
The Globe and Mail reports TD is taking a full floor, moving risk, product and platform teams in beside the lab, and raising the innovation share of its technology budget to 45% from 34% over three years.
- Apr 16, 2026
Chun on Anthropic and personal AI agents for front-line colleagues[29]
At the annual meeting he tells the Globe and Mail that TD does some work with Anthropic but is not in the limited Claude Mythos preview, and that TD plans a personal AI agent for each employee that summarises a client's portfolio and past conversations.
- Feb 26, 2026
Restructuring programme concluded: C$886 million of charges, about C$775 million of savings[23][21]
The Q1 2026 filing says the programme ended on 31 January 2026 with savings including a workforce reduction of about 3%; Chun adds that AI is helping power the C$2 billion to C$2.5 billion structural cost target through repeatable patterns.
- Feb 26, 2026
Q1 2026 call: about C$170 million of AI value in 2025[17][13]
The bank says AI delivered about $170 million in 2025; Chun says the generative knowledge-management tool has reached more than 1,000 Canadian branches and the agentic mortgage project is scaling; the CFO targets $150 million of insurance claims-cost savings.
- Jan 6, 2026
US patent 12,517,812 B2 granted: security testing with generative AI[47]
A patent assigned to The Toronto-Dominion Bank on using generative AI models to test software for security weaknesses; a continuation application was published in 2026.
- Dec 4, 2025
Q4 2025 call: about 75 AI use cases and C$170 million of value; C$200 million targeted for 2026[22]
Chun lists customer acquisition, customer insights and risk management as investment categories; fiscal 2025 fraud losses were down 26%, which the call attributes to fraud modernisation.
- Dec 4, 2025
2025 annual report: 'Simpler & Faster' technology leveraging AI[2]
The MD&A makes AI a pillar of the strategy, names AI-driven automatic adjudication in TD Auto Finance and Small Business Banking, and records the opening of a Layer 6 GenAI centre in New York and further machine learning and specialised AI in US AML remediation.
- Oct 9, 2025
TD Securities assistant: about 1,000 users, OpenAI enterprise GPT on Azure[32]
CIO Dan Bosman says OpenAI's enterprise version was a natural extension of TD's Microsoft and Azure cloud partnership; global markets rollout planned within a month; about 2,500 people build AI platforms and applications at TD.
- Oct 8, 2025
Evident AI Index: TD falls to 13th[14]
TD drops from ninth to 13th and out of the top ten on competition in innovation and talent, while adding 35% more AI-focused staff year over year.
- Sep 29, 2025
- Sep 11, 2025
OSFI publishes Guideline E-23 on model risk[19]
Canada's prudential supervisor finalises its model-risk guideline, effective 1 May 2027 for federally regulated institutions, with added clarity for AI and machine-learning models.
- Jul 8, 2025
TD Securities AI Virtual Assistant rolled out[31]
Built by TD Securities, Layer 6 and technology teams on OpenAI GPT models with retrieval and text-to-SQL; the CIO says patent-pending prompt optimisations meant the OpenAI model needed no fine-tuning.
- Jun 11, 2025
- May 22, 2025
Restructuring programme: 2% of the workforce, up to C$650 million of annual savings[37]
Reuters: about 2,000 employees, with C$600 million to C$700 million of charges expected; Chun said TD was looking for opportunities to automate and re-engineer operations and to scale digital and AI investment.
- Apr 22, 2025
Layer 6 to open a New York office[36]
An initial 20 data scientists, applied machine learning scientists and generative AI specialists at One Vanderbilt, to support the US bank and bring generative AI delivery in-house.
- May 16, 2024
Generative AI pilots: contact-centre assistant and GitHub Copilot[33]
Luke Gee, then Head of AI and Analytics, announces a Layer 6 generative assistant for contact-centre colleagues and a GitHub Copilot pilot for engineers, which followed a September to December 2023 unit-testing trial.
- Feb 22, 2024
Layer 6 published 14 papers in 2023 and had filed more than 60 patent applications[34]
American Banker: Layer 6 had about 200 people in TD's AI and machine learning team, more than 67 AI use cases across the bank and a 2023 NeurIPS award for a paper on tabular data understanding and generation.
Where does TD Bank Group use AI today?
| System | What it does | Status | Use case |
|---|---|---|---|
| Agentic pre-adjudication for mortgages and HELOCs[9][3][15] | Autonomous agents classify documents, calculate income, check policy and consent and write an underwriter summary; the Q3 2026 call said the first RESL agentic capability was fully scaled and speeding decisions, with the same approach planned for small business and auto finance. | In production | Credit scoring & underwriting |
| Automated adjudication in auto finance and small business[2][3] | The 2025 annual report plans AI to raise automatic-adjudication rates in TD Auto Finance and Small Business Banking; by August 2026 TD Auto Finance Canada had automated about a third of manual funding processes and launched digital income verification. | Rolling out | Credit scoring & underwriting |
| Generative knowledge management for frontline colleagues[17][3][16] | Introduced in contact centres in 2025, deployed across more than 1,000 Canadian branches by February 2026 and supporting over 20,000 client-facing colleagues by August 2026; answers that previously sent colleagues through multiple screens now come in seconds, per the CEO. | In production | Generative & agentic AI |
| TD AI Prism predictive foundation model[6][13] | One model in place of dozens of predictive models, used to personalise marketing; Volkovs reports 20% to 30% better predictive accuracy, 53% better lead quality and a more than 40% lift in campaign response rates. | In production | Generative & agentic AI |
| Copilot and AI-assisted engineering[4][3] | More than 40,000 colleagues use Copilot and more than 7,000 engineers use AI for software development, the most active with a 29% increase in throughput; the software development life cycle is one of three first-wave priorities. | In production | AI-generated code & coding agents |
| Contact-centre AI and collections agent[3][15] | Contact centres are a first-wave priority to simplify routine interactions; a first collections agent lifted the connect rate to 20% to 25% from 7%, per the CEO in September 2026. | Rolling out | Customer-facing chatbots |
| AI in TD Insurance claims[3][17][4] | AI-powered vehicle damage estimation is deployed at scale in Canada and the bank targets $150 million of claims-cost reductions in the medium term through AI, vendor optimisation, fraud detection and process reengineering; a client-facing generative AI virtual assistant launched in insurance in 2026. | In production | Fraud detection |
| Cohere collaboration with Layer 6[12] | Up to C$25 million over three years to apply Cohere's secure enterprise models with Layer 6 research to knowledge management, security and responsible adoption. | Announced | Third-party & vendor AI |
| Responsible AI Principles and Trustworthy AI team[10][9] | Seven commitments, model risk and privacy impact assessments, independent oversight reviews and post-deployment monitoring of fairness, performance and explainability; controls such as retrieval grounding and human oversight apply to colleague-facing assistants. | In production | AI governance (general) |
| Machine learning and specialised AI in US AML remediation[1][2] | The 2025 annual report and the Q3 2026 filing list further deployments of machine learning and specialised AI, a new KYC platform and continued investment in analytics, machine learning and AI within financial-crimes risk management as parts of the US BSA/AML programme (covered at /banks/td-bank-us). | Rolling out | AML / KYC |
| TD Securities AI Virtual Assistant[31][32] | An OpenAI GPT assistant with retrieval over TD Securities research and market data and text-to-SQL, used by about 1,000 people in October 2025 and planned for all of global markets; no fine-tuning of the OpenAI model was needed. | In production | Trading & capital markets |
| TDI Virtual Assistant (client-facing)[26] | Unauthenticated generative AI chatbot on select TD Insurance pages, built with Layer 6; it expands through 2026 in English and French. TD also reports the contact-centre assistant has cut call hold times by 12% and the Branch Virtual Assistant supports 17,000 front-line colleagues. | Rolling out | Customer-facing chatbots |
| AI collections agent[24][15] | TD's first collections agent raised the rate at which it reaches customers behind on loan payments to between 20% and 25% from about 6% to 7% with human callers, per Chun in September 2026. | Rolling out | Customer-facing chatbots |
| Agentic mortgage funding and discharge[25][24] | Two October 2026 capabilities verify documents across systems and flag discrepancies for colleague review; Chun put mortgage funding cost at about C$187 a file falling to about C$47, and discharges at about C$22 to C$23 falling to about C$11. | Rolling out | Generative & agentic AI |
| AI inside the model risk and three-lines-of-defence framework[2] | The 2025 annual report says all models, including machine learning and generative AI models, go through TD's model lifecycle process with independent validation; businesses own AI risk and Compliance, Data Risk, Model Risk, Technology and Cybersecurity Risk, Privacy and Third-Party Risk oversee it. | In production | Model risk management |
Use-case links open the tracker's rules-by-use-case matrix: which documents govern each system type.
What has TD Bank Group disclosed in numbers?
| Metric | Value | As of |
|---|---|---|
| Total assets | C$2,111.9 billion[1] | Jul 31, 2026 |
| Medium-term AI value target | C$1 billion a year (C$500 million revenue uplift, C$500 million cost savings)[5][16] | Aug 27, 2026 |
| AI value delivered, fiscal 2026 to date | C$195 million against a C$200 million full-year target[5] | Jul 31, 2026 |
| AI value delivered in 2025 | About $170 million[13][17] | Dec 31, 2025 |
| Evident AI Index 2025 rank | 13th of 50 (ninth in 2024); AI-focused staff up 35% year over year[14] | Oct 8, 2025 |
| Colleagues using Copilot | Over 40,000[4] | May 28, 2026 |
| Engineers using AI for development | Over 7,000; most active up 29% in throughput[4] | May 28, 2026 |
| Client-facing colleagues on generative knowledge management | Over 20,000[3] | Aug 27, 2026 |
| Mortgage pre-adjudication time | From an average of 15 hours to under three minutes[9] | May 21, 2026 |
| People working on AI development | Roughly 2,500[16] | Feb 20, 2026 |
| TD Insurance claims-cost target | $150 million in the medium term[17] | Feb 26, 2026 |
| Layer 6 and Cohere commitment | Up to C$25 million over three years[12] | Sep 22, 2026 |
| Restructuring programme total charges | C$886 million pre-tax, with about C$775 million of annual savings including a workforce reduction of about 3%[23] | Jan 31, 2026 |
| AI use cases implemented in fiscal 2025 | About 75, delivering C$170 million; C$200 million of incremental value targeted for fiscal 2026[22] | Dec 4, 2025 |
| Evident AI Banking Index 2026 rank | 10th of 50 (13th in 2025); measurable impact reported for 20% to 25% of use cases[27] | Oct 6, 2026 |
| Layer 6 headcount | About 90 in April 2026, growing to about 240 after the MaRS expansion[28] | Apr 18, 2026 |
| Share of technology budget for innovation and development | 34%, to rise to 45% over three years[28] | Apr 18, 2026 |
| Collections contact (connect) rate | 20% to 25% with the AI agent, against about 6% to 7% with people[24] | Sep 9, 2026 |
| Mortgage funding and discharge cost per file | Funding about C$187 to about C$47; discharge about C$22 to C$23 to about C$11 (CEO targets)[24] | Sep 9, 2026 |
| Contact-centre call hold time | Down 12% since the generative assistant launched; Branch Virtual Assistant supports 17,000 front-line colleagues[26] | May 1, 2026 |
| Share of bank AI research publications, 2024 | 4% (fifth among banks); Layer 6 had filed more than 60 patent applications by February 2024[42][34] | Sep 25, 2025 |
“Three quarters into the year, we have essentially hit our fiscal 2026 target of $200 million in value from AI.”[3]
Raymond Chun · Group President and CEO · Aug 27, 2026
“I believe AI will transform our operations, make our colleagues more efficient, our processes faster and our products and services better.”[4]
Raymond Chun · Group President and CEO · May 28, 2026
“What I am seeing with agentic AI capabilities is something that I have not seen with automation, I have not seen in digital”[15]
Raymond Chun · Group President and CEO · Sep 9, 2026
“We're building a hybrid future where our colleagues and AI work together to help our clients get to a 'yes' faster on some of their most important decisions.”[9]
Luke Gee · Chief Analytics and AI Officer · May 21, 2026
“Operating one foundation model instead of dozens of individual predictive models allows us to more efficiently implement the standards and guardrails set by our 'Trustworthy AI' team”[13]
Maksims Volkovs · SVP and Chief AI Scientist, TD · Jun 1, 2026
“suggest that the bank could well regain ground”[14]
Alexandra Mousavizadeh · Co-founder and CEO, Evident Insights · Oct 8, 2025
“A core tenet of our AI strategy is to build once and use many times – scaling AI through repeatable patterns that lead to faster AI deployments and reduce cost of delivery.”[21]
Raymond Chun · Group President and CEO · Feb 26, 2026
“When we were using people, the connect rate was about 6%, 7%. Our connect rate using our agents now with the AI behind it and getting to them at the right times, is between 20%-25%.”[24]
Raymond Chun · Group President and CEO · Sep 9, 2026
“Our job is to create an environment in which the model does what we want it to do, only what we wanted to do, and frankly, it does it really bloody well.”[30]
Chad Koziel · Assistant Vice President, TD, and founder of Layer 6 · May 28, 2026
“One thing we've been really clear on is making sure this thing is not hallucinating.”[32]
Dan Bosman · Chief Information Officer, TD Securities · Oct 9, 2025
“we fully scaled our first RESL agentic AI capability, and it's absolutely delivering faster decisions.”[3]
Sona Mehta · Group Head, Canadian Personal Banking · Aug 27, 2026
Who runs AI at TD Bank Group?
Luke Gee
Former Senior Vice President, Chief Analytics and AI Officer (title in TD releases to 23 July 2026; Head of AI and Analytics in May 2024); left for Scotiabank, whose 21 October 2026 statement names him Executive Vice-President and Chief Data and AI Officer[9][8][33][7]
Paul Whitehead
Senior Executive Vice President, Global Corporate Services; responsible for enterprise AI (from 2026-09-04)[11]
Sumee Seetharaman
Head of AI and ML Practice; quoted on the Responsible AI Principles (July 2026); full title Head of Artificial Intelligence and Machine Learning Practice[10]
Ajai Bambawale
Chief Risk Officer; quoted on the Responsible AI Principles (July 2026)[10]
Rizwan Khalfan
Executive Vice President and Vice Chair, Strategic Growth, Innovation and Partnerships[12]
Mohit Veoli
Senior Vice President, Real Estate Secured Lending[9]
Taylan Turan
Former Group Head and Chief Operating Officer (leaving TD, announced 2026-09-03)[11][18]
Licenia Rojas
Senior Vice President, Chief Engineer and Architect (as of May 2024; sponsor of the GitHub Copilot pilot for engineers)[33]
Jesse Cresswell
Staff Machine Learning Scientist, Layer 6; leads the Trustworthy AI team; inventor on US 12,639,628 B2[30][49]
Chad Koziel
Assistant Vice President, TD, and founder of Layer 6; led the first agentic AI build[30]
Dan Bosman
Chief Information Officer, TD Securities (built the TD Securities AI Virtual Assistant with Layer 6)[32][31]
Ted Paris
Senior Vice President, Head of Analytics, Intelligence and AI, TD Bank U.S.[38]
Kiran Vuppu
U.S. Chief Information Officer, TD Bank[39]
Eric Huber
Head of Adversarial Intelligence and Disruption, TD[40]
Jo Jagadish
Head of digital banking and contact centers, TD (title in November 2025; led the contact-centre AI knowledge tool rollout)[41]
Sona Mehta
Group Head, Canadian Personal Banking (business owner of the first RESL agentic AI capability)[3]
What AI roles is TD Bank Group hiring for?
Data as of · 12 AI-titled postings · read from TD Bank Group's own careers site, re-checked weekly
12 open AI-related postings on TD Bank Group's careers site as of Nov 10, 2026. A snapshot of current openings, not a source: postings close. Compare with other banks on which banks are hiring for AI.
Director, AI Governance, TDS Governance & Control
New York, New York
Toronto, Ontario · posted
Toronto, Ontario · posted
Toronto, Ontario · posted
Technical Product Owner - AI & Machine Learning Solutions
Toronto, Ontario
Senior Applied Machine Learning Scientist, Generative AI Model Validation ATH# 5480
Montral Qubec (+1 more) · posted
Senior Machine Learning Engineer
Toronto, Ontario
Research Machine Learning Scientist
Toronto, Ontario
Mount Laurel New Jersey (+1 more) · posted
Data Scientist III - FCRM Risk Modeling
Mount Laurel New Jersey (+4 more) · posted
Applied Machine Learning Scientist I ATH# 5645 (Tabular Deep Learning Model Validation)
Toronto, Ontario · posted
Engineer II - Senior Agentic AI IGA Developer
Toronto, Ontario · posted
Which regulators govern TD Bank Group's AI?
| Authority | Why it matters here | Documents |
|---|---|---|
| OSFI | TD's home prudential supervisor. Guideline E-23 on model risk management takes effect for federally regulated institutions on 1 May 2027 and adds clarity for AI and machine-learning models; OSFI's July 2026 bulletin sets sound practices for generative and agentic AI, directly relevant to TD's agentic lending model. | |
| OCC | TD Bank, N.A., the group's US bank, is a national bank under an OCC consent order; group AI platforms used there are examined inside that remediation (covered at /banks/td-bank-us). | OCC Bulletin 2026-13 OCC Bulletin 2023-17 |
| Federal Reserve | TD's US holding company is Federal Reserve supervised, so model-risk guidance such as SR 26-2 reaches AI models built in Toronto and used by US entities. | SR 26-2 SR 23-4 |
Side by side: SR 11-7 vs SR 26-2 · OCC vs CFPB on AI lending · PRA SS1/23 vs SR 26-2 · MAS AI Risk Management Guidelines vs SR 26-2 · all comparisons.
What does the public record suggest about TD Bank Group's AI strategy?
Analysis by BankingNewsAI from the sources cited on this page. Observations, not advice.
01
A dollar target reported every quarter
Few banks give a running figure for AI value. TD reports it against a fiscal-year target (C$195 million of C$200 million by Q3 2026) and a medium-term C$1 billion goal, which makes slippage or upside visible, and Chun has already said he may have underestimated the upside.
02
A lab-first bank that fell in the rankings and recovered
TD has had Layer 6 since 2018 and fell from ninth to 13th on Evident's 2025 index as peers moved faster on innovation and talent, then returned to 10th in 2026 after reporting C$195 million of AI value in nine months. Evident said 20% to 25% of TD's use cases report a measurable impact, so the open question is how many of the rest do.
03
Agents that can be reused
The mortgage agent is being extended to small business and auto finance, and the CEO's 'build once and use many times' principle points to a platform pattern rather than a series of one-off tools. Reuse also concentrates model risk in the shared components.
04
Accountability moved to a corporate-services executive
Placing enterprise AI under Global Corporate Services in September 2026, alongside transformation and marketing, while technology stays with Global Technology and Solutions, splits delivery from enterprise direction.
Where does the TD Bank Group record go deeper?
3 topic pages under this profile, each answering one question from the same sourced record.
Which leader rankings does TD Bank Group appear in?
45 of the computed leaders rankings include TD Bank Group, each ranked from the same sourced records (bank pages, filings, hiring snapshots, the agent tracker, the incidents register).
How does TD Bank Group compare with its peers?
- How does The Toronto-Dominion Bank (TD Bank Group) compare with its peers on AI and technology?
- The Toronto-Dominion Bank (TD Bank Group) competitors and alternatives, by business line
Its peer group on this site, by business model and size:
What is TD's AI value target?
TD aims to generate C$1 billion a year of value from AI over the medium term, half from cost savings and half from revenue, a target Chun set at the September 2025 Investor Day. By the third quarter of fiscal 2026 it had delivered C$195 million against a C$200 million target for the year.
What is Layer 6?
Layer 6 is the Toronto AI company TD acquired in 2018. It is the bank's AI research and development centre, with offices in Toronto and New York, and built TD AI Prism, a causal foundation model and the bank's first agentic AI model.
How does TD rank for AI?
TD ranked 10th of 50 banks in the 2026 Evident AI Banking Index published on 6 October 2026, up from 13th in 2025 and ninth in 2024, according to the Financial Post's reports of the index. Royal Bank of Canada ranked third in 2025 and 2026; its record is at /banks/rbc.
Does TD use AI in the United States?
Yes. TD Bank, N.A. is covered separately at /banks/td-bank-us, including AI in anti-money-laundering monitoring under the OCC consent order. The group's Layer 6 opened a New York office to support US demand.
Who is responsible for AI at TD?
Maksims Volkovs is Chief AI Scientist at Layer 6. Luke Gee, Chief Analytics and AI Officer in TD's releases to July 2026, has left for Scotiabank, where he is Chief Data and AI Officer. From 4 September 2026 Paul Whitehead, Senior Executive Vice President, Global Corporate Services, is responsible for enterprise AI.
Who leads AI at TD Bank Group?
Direction, delivery and research sit with different people. CEO Raymond Chun owns the C$1 billion value target. From 4 September 2026 Paul Whitehead, Senior Executive Vice President, Global Corporate Services, is responsible for enterprise AI. Vlad Shpilsky runs Global Technology and Solutions, which Fintech Futures says covers AI strategy, data, fraud operations and cyber resilience. Maksims Volkovs is Chief AI Scientist at Layer 6, the research lab, and Sumee Seetharaman heads the AI and machine learning practice. Luke Gee, Chief Analytics and AI Officer in TD's releases to July 2026, has since left for Scotiabank as its Chief Data and AI Officer. Ajai Bambawale, the Chief Risk Officer, is the risk executive named on the Responsible AI Principles.
Which AI vendors and partners does TD Bank Group use?
TD builds core models at Layer 6 and buys selectively. It piloted GitHub Copilot, a Microsoft product, with engineers from September 2023. The TD Securities assistant runs on OpenAI enterprise GPT models, chosen as an extension of TD's Microsoft Azure relationship. In September 2026 Layer 6 and Cohere agreed up to C$25 million over three years. Chun said in April 2026 that TD does some work with Anthropic but was not in the limited Claude Mythos preview. TD also joined the MIT Media Lab's sAIpien programme in October 2025.
Is TD cutting jobs because of AI?
TD has not tied job cuts to AI in its filings. A restructuring programme announced in May 2025 covered about 2,000 employees, 2% of the workforce, and concluded on 31 January 2026 with C$886 million of pre-tax charges and about C$775 million of annual savings, including a workforce reduction of about 3%. The filing lists severance, real estate and business wind-downs. At the launch Chun said TD was looking to automate and re-engineer operations, and TD names AI as a lever for its C$2 billion to C$2.5 billion cost target. It is also adding 1,000 branch staff and 1,200 wealth advisers.
- TD Bank Group Third Quarter 2026 Report to Shareholders (Form 6-K, Exhibit 99.1)
- TD Bank Group 2025 Annual Report: Management's Discussion and Analysis (Form 40-F, Exhibit 99.2)
- TD (TD) Q3 2026 Earnings Call Transcript
- TD (TD) Q2 2026 Earnings Call Transcript
- TD Bank Unlocks $141 Million in AI Value Months Ahead of Schedule
- TD Announces Launch of Groundbreaking Predictive Foundation Model
- Scotiabank hires new AI officer from Wells Fargo
- Canadians feel pressure to keep up with AI at work, new TD survey finds
- TD Launches Agentic AI to Transform Real Estate Secured Lending from End to End
- TD strengthens Its Responsible AI Principles to support trust in AI innovation
- TD Bank Group announces senior leadership appointments to further accelerate strategy
- TD announces $25 million strategic relationship with Cohere to support AI adoption
- #4 Maksims Volkovs drives TD Bank's AI transformation
- RBC retains top spot, TD stumbles as AI race in banking intensifies
- Banks say AI is allowing them to recover unpaid loans, cut call centre volumes and save workdays
- How AI is Permeating Work Throughout TD Bank, and Lessons Learned
- TD eyes $150M in cost reductions – with help from AI
- TD wants to get 'back to winning'
- Guideline E-23 – Model Risk Management (2027) - Letter
- Generative and Agentic Artificial Intelligence: Implications for Technology, Cyber Security, and Operational Resilience
- TD Bank Group Q1 2026 Earnings Conference Call transcript
- TD Bank Group Q4 2025 Earnings Conference Call transcript
- TD Bank Group First Quarter 2026 Report to Shareholders (Form 6-K, Exhibit 99.1)
- TD Bank at Scotiabank summit: profit gains, AI push lift outlook (transcript)
- TD expands agentic AI across mortgage operations to make banking faster, simpler and more human
- Let's Chat Coverage: TD Insurance Rolls Out New Virtual Assistant
- TD's AI returns help put it back into the top 10 in this global bank index
- How the promise of AI is taking hold at Canada's biggest banks
- Anthropic taking responsible approach with Claude Mythos software rollout, TD CEO says
- TD's new AI agent shaves 15 hours off mortgage decisions
- TD Securities taps Layer 6 and OpenAI to deliver real-time equity insights to sales and trading teams
- 'Give me five trade ideas': TD's traders embrace ChatGPT
- TD launches new generative AI pilots to help empower colleagues to deliver legendary customer experiences
- JPMorgan Chase, TD draw AI talent through research labs
- TD joins the Massachusetts Institute of Technology's Media Lab to Explore AI Innovation in Financial Services
- TD Announces Opening of New Layer 6 Office in New York City
- Canada's TD Bank to lay off 2% of workforce in restructuring
- TD's US head of AI weighs in on agents
- Banks batten down the hatches for rogue AI agent swarms
- $300 AI tool kits let criminals bypass bank security
- How Citi, Citizens, TD deploy AI agents without upsetting humans
- Financial firms fuel a surge in AI research, adoption
- US 12,079,351 B2: Application of trained artificial intelligence processes to encrypted data within a distributed computing environment
- US 12,321,861 B2: Initialization of parameters for machine-learned transformer neural network architectures
- US 12,326,856 B2: Using generative artificial intelligence to improve user interactions
- US 12,399,687 B2: Generating software architecture from conversation
- US 12,517,812 B2: Security testing based on generative artificial intelligence
- US 12,592,301 B2: Prompt engineering and generative AI for goal-based imagery
- US 12,639,628 B2: Distributed model training with collaboration weights for private data sets
- US 12,657,566 B2: GenAI letter of confirmation of benefits
- US 2025/0077397 A1: Large language model training for test case generation
- US 2025/0119396 A1: Charge card knowledge chatbot
When TD Bank Group 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 PT · free
plus every regulation, bank and agent page update, the morning after · leave any morning