An in-house lab (Layer 6), a proprietary foundation model, and a dollar target for AI value — under a US asset cap that makes efficiency the growth plan.
How is TD using AI?
TD's AI strategy is set at group level in Toronto and executed on both sides of the border: a 'pragmatic buy, build and blend' approach in which Layer 6, the bank's own AI lab, builds where TD has a data advantage — the TD AI Prism foundation model, generative assistants for branch, wealth and markets staff, and its first agentic model, which cuts hours from mortgage pre-adjudication — and buys where speed matters. The program has a dollar target: C$200 million of AI value in fiscal 2026 on the way to C$1 billion a year, and the bank said in August 2026 it had essentially hit the year's target three quarters in. For the US bank, operating under the asset cap imposed after the 2024 anti-money-laundering settlement, AI is the growth lever that does not need a bigger balance sheet, and the AML rebuild is where much of it is aimed. US head Leo Salom told analysts in May 2026 that the US AML program 'is now running on a new transaction monitoring system that has an embedded machine learning and AI enhancements', and in September 2026 said three AI models are deployed in the AML program (including adverse-news scanning and unusual-transaction detection), 1,800 US developers have GitHub Copilot with technology-capacity gains of 30% to 40%, and a vendor voice AI agent is targeted to handle about half of incoming call-center calls. The group's 2025 annual report, filed with the SEC, lists 'further deployments of machine learning and specialized AI' in the US BSA/AML remediation and names the New York Layer 6 GenAI centre as a US Retail achievement.
Lead bank
TD Bank, N.A. (TD Group US Holdings)
Assets (lead bank, Mar 31, 2026)
$346 billion
Charter
National bank (OCC)
Headquarters
Cherry Hill, NJ (TD Bank Group: Toronto)
Flagship platform
Layer 6 and TD AI Prism
AI leadership
Maksims Volkovs · Ted Paris · Raymond Chun · Leo Salom · Vlad Shpilsky · Kiran Vuppu · Jo Jagadish
- Layer 6, acquired in 2018, is the in-house AI R&D centre; a New York office opened in 2025 to support US demand; chief AI scientist Maksims Volkovs has overseen nearly 100 use cases and 150-plus patents.
- TD AI Prism (June 2025): a predictive foundation model built in-house over three years to anticipate customer needs; an American Banker Innovation of the Year 2026 honoree.
- Generative knowledge-management assistants: TD Securities (July 2025, pilot), TD Wealth (October 2025), branches (national expansion, Q1 fiscal 2026), TD Insurance (May 2026).
- May 2026: first agentic AI model, generating application summary memos for real-estate secured lending underwriters in minutes; American Banker reports 15 hours shaved off mortgage decisions.
- Value target: C$200 million of AI value in fiscal 2026, essentially met by the third quarter; C$1 billion a year of enterprise value from AI set out at the September 2025 investor day; $150 million of insurance claims cost reduction targeted.
- US AML: a new transaction monitoring system with embedded machine learning and AI, three AI models in the AML program and an investigator desktop in development (May and September 2026).
- US productivity: 1,800 developers on GitHub Copilot with 30-40% technology-capacity gains; a voice AI agent targeted at about 50% of call-center calls; group-wide, over 40,000 colleagues use Copilot (vendor not named) and 7,000-plus engineers use AI for software development.
- Open AI roles include a head of mobile and agentic banking transformation (US), a senior gen AI / agentic AI engineer (US) and US data scientists for financial-crimes modelling (listed below, as of 24 September 2026).
- Governance: enterprise Responsible AI Principles (July 2026) with seven commitments, an AI Risk Management framework and a Trustworthy AI team; a June 2026 Reuters report on workforce-monitoring software in financial-crimes teams shows the friction.
What is Layer 6 and TD AI Prism?
Layer 6, the AI research centre TD acquired in 2018, builds the group's models in-house: TD AI Prism, a predictive foundation model launched in June 2025 after three years of R&D; generative knowledge-management assistants for branches, wealth and TD Securities; and the bank's first agentic model, for real-estate secured lending, in May 2026. Chief AI scientist Maksims Volkovs' 130-plus team has overseen nearly 100 use cases and filed more than 150 patents.[1][2][7][10]
What has TD done on AI, and when?
- Sep 22, 2026
US CIO on guardrails against 'rogue' agentic AI swarms[33]
Kiran Vuppu tells American Banker that agents given a goal can take many possible routes to it, so banks cannot predict every scenario and need continuous monitoring of agent reasoning (chain-of-thought observability) alongside guardrails and human oversight at both design and operational stages.
- Sep 22, 2026
Layer 6 and Cohere: up to C$25 million over three years[31]
Strategic collaboration on enterprise AI, with a focus on knowledge management, security and responsible adoption.
- Sep 17, 2026
US bank details AI in AML, engineering and the call centre[30]
Three AI models in the AML program, an investigator desktop in development, 1,800 developers on GitHub Copilot, and a vendor voice AI agent targeted at about 50% of incoming calls.
- Sep 3, 2026
Shpilsky reappointed to lead Global Technology and Solutions[32]
Senior leadership appointment continuing his mandate over the group's technology and digital organisation, which has 'deployed new innovations and platforms and launched new AI-enabled applications'.
- Aug 27, 2026
- Aug 27, 2026
- Jul 30, 2026
Responsible AI Principles published[3]
Seven commitments embedded in the AI Risk framework, overseen by AI Risk Management and a Trustworthy AI team.
- Jun 19, 2026
Work-monitoring software in financial-crimes teams[19]
Reuters: tracking time in browsers and chat apps prompts consent and privacy questions.
- May 28, 2026
Q2 fiscal 2026 call: US AML runs on AI-enhanced transaction monitoring[27]
Salom: new monitoring system with embedded ML and AI, new KYC platform and improved customer risk rating model; group: over 40,000 colleagues using Copilot (the call did not name the product's vendor), over 7,000 engineers using AI, mortgage pre-adjudication cut from about 15 hours to 3 minutes.
- May 21, 2026
- Mar 31, 2026
Consumer survey: 55% use AI for money decisions, up from 10%[13]
TD's second AI Insights survey of 2,500 US consumers.
- Feb 25, 2026
- Feb 18, 2026
US head of AI on agents and failure rates[14]
Ted Paris: agents can 'truly transform operations' but must be fitted to the bank's environment.
- Dec 4, 2025
2025 annual report (SEC filing): AI risk framework and US AML machine learning[25]
AI models held to the same model-risk standards; generative and agentic AI named as risks; 'further deployments of machine learning and specialized AI' in the US BSA/AML remediation; ~US$52 billion of headroom to the US asset limitation.
- Oct 20, 2025
TD Wealth Virtual Assistant[4]
A generative knowledge-management system built by Layer 6 for wealth colleagues.
- Sep 29, 2025
- Jul 8, 2025
TD Securities AI Virtual Assistant pilot[5]
A proprietary generative chatbot for front-office sales, trading and research.
- Jun 11, 2025
TD AI Prism launched[1]
A predictive foundation model to anticipate customer needs, built entirely in-house.
- May 22, 2025
2% of workforce cut to fund digital and AI[18]
About 2,000 roles in a restructuring after the US AML settlement, under new CEO Ray Chun.
- Apr 22, 2025
Layer 6 opens a New York office[6]
To bring gen-AI expertise and delivery in-house for the US bank, per Leo Salom.
- Jan 1, 2018
TD acquires Layer 6[6]
The Toronto AI research firm becomes the group's AI R&D centre.
Where does TD use AI today?
| System | What it does | Status | Use case |
|---|---|---|---|
| Agentic lending pre-adjudication[2][12] | Application summary memos for real-estate secured lending underwriters. | Rolling out | Credit scoring & underwriting |
| TD AI Prism personalisation[1][11] | Predictive foundation model for customer needs and marketing. | In production | Customer-facing chatbots |
| Knowledge-management assistants[4][5][8] | Branch, wealth, securities and insurance virtual assistants built by Layer 6. | In production | Generative & agentic AI |
| Fraud defence and claims[15][16] | AI in fraud defence and insurance-claims fraud detection; $150 million claims-cost target. | In production | Fraud detection |
| AML rebuild[22][19] | The US AML program rebuild after the 2024 settlement, with technology and monitoring investment. | Rolling out | AML / KYC |
| AI models in US AML monitoring and investigations[27][30][25] | New transaction monitoring system with embedded ML and AI; three AI models including adverse-news scanning and unusual-transaction detection; investigator desktop in development. | In production | AML / KYC |
| Voice AI agent for US contact centres[30] | Built with a vendor; targeted to handle about 50% of incoming call-center calls. | Announced | Customer-facing chatbots |
| AI-assisted software development[30][27] | 1,800 US developers on GitHub Copilot with 30-40% technology-capacity gains; 7,000-plus engineers group-wide, the most active achieving a 29% increase in throughput. | In production | Generative & agentic AI |
| AI risk management under the model lifecycle[25] | Annual report: AI models (including ML and generative AI) meet the same model-risk standards; a three-lines-of-defense approach to AI risk; generative and agentic AI adoption by TD and its vendors named as a risk. | In production | Model risk management |
| Agent observability and guardrails[33] | US CIO describes chain-of-thought monitoring of agent reasoning plus guardrails and human oversight at design and operational stages, to contain agents that find unpredictable routes to a goal. | In production | AI governance (general) |
| Automated bond trading[20] | A computer-driven trading team wins more US investment-grade business with fewer people. | In production | Trading & capital markets |
Use-case links open the tracker's rules-by-use-case matrix: which documents govern each system type.
Which AI vendors and fintechs does TD work with?
| Company | What TD uses it for | Status | Category |
|---|---|---|---|
| FICO | FICO Platform in the TD North American Fraud Detection Platform (a TD group platform covering its North American operations, including the US) | In production | Credit decisioning |
| GitHub Copilot | 1,800 developers enabled with GitHub Copilot, with technology-capacity improvements of 30% to 40% depending on adoption. | In production | Enterprise AI assistants |
Each vendor and fintech page cites the source for the relationship, and vendor pages list products, pricing and other bank customers. Compare AI tools for banks by category, or read the fintech and neobank AI dossiers.
What has TD disclosed in numbers?
| Metric | Value | As of |
|---|---|---|
| AI value target | C$200 million (FY2026), C$1 billion a year medium term[23][17] | Aug 27, 2026 |
| AI value delivered | C$195 million (about US$141 million) in the first three quarters of FY2026; almost C$145 million at the half-year[23][27] | Jul 31, 2026 |
| AI tools in use (group) | Over 40,000 colleagues using Copilot (vendor not named); over 7,000 engineers using AI for software development[27] | May 28, 2026 |
| US developers on GitHub Copilot / technology-capacity gain | 1,800 / 30-40%[30] | Sep 17, 2026 |
| Headroom to the US asset limitation | ~US$52 billion after a 10% asset reduction[25] | Oct 31, 2025 |
| Layer 6 use cases / patents | ≈ 100 / 150+[7] | Jan 16, 2026 |
| Layer 6 team | 130+[10] | Jun 1, 2026 |
| Claims-cost reduction targeted | $150 million (medium term)[16] | Feb 26, 2026 |
| Workforce cut to fund AI/digital | ≈ 2% (≈ 2,000 roles)[18] | May 22, 2025 |
“Three-quarters into the year, we have essentially hit our fiscal 2026 target of 200 million [Canadian dollars] in value from AI.”[23]
Raymond Chun · Group President and CEO · Aug 27, 2026
“AI agents can truly transform operations.”[14]
Ted Paris · Head of Analytics, Intelligence and AI, TD Bank U.S. · Feb 18, 2026
“As you've heard me speak about in prior quarters, our AML program is now running on a new transaction monitoring system that has an embedded machine learning and AI enhancements.”[27]
Leo Salom · President and CEO, TD Bank (U.S.) · May 28, 2026
“Within financial services, our responsibility is to engineer AI solutions where those safeguards are consistent and enforced.”[26]
Kiran Vuppu · U.S. Chief Information Officer, TD · Mar 31, 2026
“When you peel back the onion, it really comes down to ensuring you have the right guardrails and observability in place.”[33]
Kiran Vuppu · U.S. Chief Information Officer, TD · Sep 22, 2026
Who runs AI at TD?
Maksims VolkovsLinkedIn ↗
Ted ParisLinkedIn ↗
SVP, Head of Analytics, Intelligence and AI, TD Bank U.S.[14][15]
Raymond ChunLinkedIn ↗
Leo SalomLinkedIn ↗
President and CEO, TD Bank (U.S.)[6]
Vlad Shpilsky
Group Head, Global Technology and Solutions, TD Bank Group[32]
Kiran Vuppu
U.S. Chief Information Officer, TD[26]
Jo JagadishLinkedIn ↗
Head of Digital Banking, Payments and Contact Centers, TD Bank U.S.[26]
What AI roles is TD hiring for?
12 open AI-related postings on TD's careers site as of Oct 1, 2026. A snapshot of current openings, not a source: postings close.
Head of Mobile and Agentic Banking Transformation (US)
Mount Laurel, New Jersey · posted Sep 29, 2026
Director, AI Governance, TDS Governance & Control
New York, New York · posted Sep 30, 2026
Senior Engineer – Gen AI /Agentic AI (US)
Mount Laurel, New Jersey · posted Sep 23, 2026
Southfield, Michigan · posted Sep 21, 2026
Senior Manager, Data Science - FCRM Modeling
New York, New York · posted Sep 25, 2026
Senior Manager, Data Science (US)
Charlotte, North Carolina · posted Sep 29, 2026
Charlotte, North Carolina · posted Sep 21, 2026
Charlotte, North Carolina · posted Sep 28, 2026
Mount Laurel, New Jersey · posted Sep 25, 2026
2027 Summer Internship Graduate Leadership Program - Retail Banking Strategy & AI
New York, New York · posted Sep 1, 2026
Applied Machine Learning Scientist II
Toronto, Ontario · posted Sep 22, 2026
Applied Machine Learning Scientist I
Toronto, Ontario · posted Sep 29, 2026
Which regulators govern TD's AI?
| Authority | Why it matters here | Documents |
|---|---|---|
| OCC | TD Bank N.A. is a national bank operating under the asset cap and AML consent order that followed the 2024 settlement; AI in monitoring is examined against that order. | OCC Bulletin 2026-13 OCC Bulletin 2023-17 |
| FinCEN | The $3 billion-plus AML penalty makes FinCEN's program rule and its 'effective use of AI' language directly relevant. | 2026 AML/CFT Program Proposed Rule 2021 BSA/AML Model Risk Management Statement |
| CFPB | Agentic lending memos feed credit decisions on natural persons; adverse-action duties apply. | ECOA / Regulation B adverse action (15 U.S.C. 1691(d); 12 CFR 1002.9) |
| Federal Reserve | The US holding company is Fed-supervised; SR 26-2 governs Prism-style models. | SR 26-2 |
| FSB | Group-level Responsible AI Principles map to the FSB's governance practices. | FSB AI sound practices consultation (June 2026) |
Side by side: SR 11-7 vs SR 26-2 · OCC vs CFPB on AI lending · PRA SS1/23 vs SR 26-2 · all comparisons.
What does the public record suggest about TD's AI strategy?
Analysis by BankingNewsAI from the sources cited on this page. Observations, not advice.
01
AI is the growth plan a capped bank can execute
With US retail assets capped, TD's value has to come from efficiency and fee businesses. A dollar-denominated AI target reported quarterly is how it shows that to investors — few peers publish one.
02
The AML rebuild is the use case that matters most to the regulator
TD's most consequential AI deployment is not customer-facing; it is in transaction monitoring and investigations under a consent order. The monitoring-software episode shows how closely that workforce is being managed.
03
Build where the data advantage is
Prism and the lending agent are trained on TD's own data; the branch and wealth assistants are knowledge tools. The 'buy, build and blend' rule is a template for banks with an in-house lab.
04
Published principles arrive after the agents, not before
Responsible AI Principles in July 2026 follow the first agentic model in May 2026. Expect supervisors to ask which principles were applied to that model at launch.
- Canada’s TD Bank pledges C$25m for AI
- TD Bank announces $25M collaboration with Cohere to facilitate AI adoption
- TD Bank Group shakes up leadership to consolidate tech, AI, and commercial strategy
- TD Bank Group Reshuffles Senior Leadership Team to Drive AI and Digital Strategy
- TD Bank Unlocks $141 Million in AI Value Months Ahead of Schedule
What is Layer 6 at TD?
Layer 6 is TD Bank Group's AI research and development centre, acquired in 2018 and based in Toronto's MaRS district with a New York office opened in 2025. Led by chief AI scientist Maksims Volkovs, its 130-plus team built the TD AI Prism foundation model, the bank's generative knowledge-management assistants and its first agentic AI model, and has filed more than 150 patents.
How much value does TD say it gets from AI?
TD set a target of C$200 million of value from AI in fiscal 2026, on the way to C$1 billion a year set out at its September 2025 investor day. On the August 27, 2026 earnings call CEO Raymond Chun said the bank had essentially hit the fiscal-year target three quarters in, across predictive, generative and agentic use cases. PYMNTS put the three-quarter figure at C$195 million (about US$141 million).
- TD Announces Launch of Groundbreaking Predictive Foundation Model
- 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
- New TD Wealth Virtual Assistant Set to Enhance the Colleague and Client Experience
- Virtual AI Assistant To Help Power TD Securities
- TD Announces Opening of New Layer 6 Office in New York City
- TD wins BIG Innovation Awards for fifth consecutive year
- TD Bank Group Reports First Quarter 2026 Results (report to shareholders)
- TD Bank Group Reports Third Quarter 2026 Results
- Maksims Volkovs drives TD Bank's AI transformation
- TD Bank Builds AI Model to Predict What Customers Need and Not What They Don't Want
- TD's new AI agent shaves 15 hours off mortgage decisions
- More Americans asking AI for financial advice: TD survey
- TD's US head of AI weighs in on agents
- TD exec keeps sight of customer trust amid AI rush
- TD eyes $150M in cost reductions – with help from AI
- TD wants to get 'back to winning'
- Canada's TD Bank to lay off 2% of workforce in restructuring
- Canadian lender TD tells some employees it will use software to monitor their work
- TD Bank Uses Automation to Trade More Bonds With Fewer People
- TD Bank Brings Back Financial Targets, Looks to Return Billions to Shareholders
- TD Bank Pushes Up Dividend After Underlying Earnings, Revenue Beat Expectations
- TD Bank Unlocks $141 Million in AI Value Months Ahead of Schedule
- TD Bank eyes $150M in claims cost reductions — with help from AI
- TD Bank Group 2025 Annual Report: Management's Discussion and Analysis (Form 40-F, Exhibit 99.2)
- Nearly 80% of Americans use AI Tools but Most Still Want Humans Making Financial Decisions, TD Survey Finds
- TD (TD) Q2 2026 Earnings Call Transcript
- TD Bank Group Third Quarter 2026 Report to Shareholders (Form 6-K, Exhibit 99.1)
- TD (TD) Q3 2026 Earnings Call Transcript
- TD Bank Targets 100 U.S. Branches as AML Remediation Continues
- TD announces $25 million strategic relationship with Cohere to support AI adoption
- TD Bank Group announces senior leadership appointments to further accelerate strategy
- Banks batten down the hatches for rogue AI agent swarms
When TD 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