How does TD Bank Group govern AI responsibly?
TD governs AI through Responsible AI Principles introduced on July 30, 2026: seven commitments covering applicable laws, transparency and explainability, data use and privacy, fairness, quality and accountability, reliability, and security. They are embedded in the AI Risk framework and policy, overseen by an AI Risk Management team and operationalized by the Trustworthy AI team, a group of scientists inside Layer 6 that evaluates use cases for fairness, explainability, privacy and performance before deployment and monitors them afterward. TD Stories says that team, led by Staff Machine Learning Scientist Jesse Cresswell, had eight people, and that its approach was named Best Responsible AI Program in the Americas at the DataIQ awards in May 2025. Controls for colleague assistants include retrieval grounding, human oversight and continuous monitoring, according to Chief Risk Officer Ajai Bambawale. In the 2026 Evident AI Index TD ranked No. 4 for responsible AI and No. 5 for AI transparency, TD says. Its supervisor OSFI finalized model-risk Guideline E-23 on September 11, 2025, effective May 1, 2027.[1][2][3][4][5]
Data as of Nov 17, 2026 · TD Bank Group profile updated Nov 8, 2026 · 17 sources cited on this page · 141 banks profiled on this site · Markdown
What are the parts of TD's responsible AI framework?
| Item | What is on the record | Source | Date |
|---|---|---|---|
| Responsible AI Principles | Seven commitments; every AI use case must align with them and undergo assessments for explainability, fairness and performance | TD release | July 30, 2026[1] |
| AI Risk Management team | Oversees the principles, which are embedded in the AI Risk framework and policy, with model risk assessments, privacy impact assessments and independent oversight reviews | TD release | July 30, 2026[1] |
| Trustworthy AI team | Within Layer 6; led by Jesse Cresswell, 'Leading a team of eight people'; human-in-the-loop review of AI recommendations | TD Stories | Undated (retrieved November 17, 2026)[2] |
| DataIQ award | TD's Trustworthy AI approach named Best Responsible AI Program in the Americas, across all industries | TD Stories | May 2025[3] |
| TD AI Prism | Built on Trustworthy AI considerations from the start; Volkovs said one foundation model makes it easier to apply the team's guardrails | TD Stories; American Banker | June 1, 2026[3][6] |
| Agentic mortgage model | Developed with the Trustworthy AI approach, including checks and balances every step of the way and a human-in-the-loop | TD Stories | Undated (retrieved November 17, 2026)[7] |
| Credit Assist | Required controls and professional reviews remain in place; credit professionals stay responsible for lending recommendations and final decisions | TD Bank U.S. | October 22, 2026[8] |
| Contact-centre assistant | Instructed to say 'I don't know' if it cannot find the information; agents rate answers with a thumbs up or thumbs down | Financial Post | June 3, 2024[9] |
| Vector training | Programs on 'explainability in large language models (LLMs)' and 'practical AI governance' attended by more than 500 TD colleagues | TD Stories | Undated (retrieved November 17, 2026)[10] |
Which rules and benchmarks apply to TD's AI?
OSFI's Guideline E-23 on model risk management, finalized on September 11, 2025, takes effect on May 1, 2027 and adds clarity for AI and machine-learning models; OSFI's July 16, 2026 bulletin sets sound practices for generative and agentic AI.[5][11]
TD's 2025 annual report says all models, including machine learning and generative AI models, go through its model lifecycle process with independent validation.[12]
TD says it ranked No. 4 for responsible AI and No. 5 for AI transparency in the 2026 Evident AI Index; Sumee Seetharaman, Head of Artificial Intelligence and Machine Learning Practice, said 'At TD, AI innovation and responsible AI are not separate conversations'.[4]
Not on the record: the full text of the seven principles as a standalone document, or the number of use cases rejected at review.[1]
What does the TD Bank Group profile record on this topic?
The rows below are the entries of the TD Bank AI strategy profile that bear on this page, as of Nov 8, 2026; the profile holds the full record of 28 dated moves, 15 use cases and 52 sources.
| Use case | What the record shows | Status | Source |
|---|---|---|---|
| Cohere collaboration with Layer 6 | 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 | [13] |
| Responsible AI Principles and Trustworthy AI team | 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 | [1][14] |
| Machine learning and specialised AI in US AML remediation | 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 | [15][12] |
| AI inside the model risk and three-lines-of-defence framework | 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 | [12] |
| Date | Move | Detail | Source |
|---|---|---|---|
| Sep 22, 2026 | Layer 6 and Cohere: up to C$25 million over three years | A strategic collaboration on knowledge management, security and responsible adoption, with a dedicated Cohere team working alongside Layer 6 at MaRS in Toronto. | [13] |
| Jul 30, 2026 | Responsible AI Principles introduced | 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. | [1] |
| Jul 16, 2026 | OSFI bulletin on generative and agentic AI | 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. | [11] |
| May 21, 2026 | First agentic AI model launched in mortgages and HELOCs | 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. | [14] |
| Apr 18, 2026 | Layer 6 to grow from about 90 staff to about 240 at MaRS | 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. | [16] |
| Dec 4, 2025 | Q4 2025 call: about 75 AI use cases and C$170 million of value; C$200 million targeted for 2026 | 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. | [17] |
| Sep 11, 2025 | OSFI publishes Guideline E-23 on model risk | 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. | [5] |
Does TD have responsible AI principles?
Yes. TD introduced seven enterprise-wide Responsible AI Principles on July 30, 2026, embedded in its AI Risk framework and policy. Every AI use case must align with them and undergo assessments for explainability, fairness and performance.
What is TD's Trustworthy AI team?
A team of scientists inside Layer 6, led by Jesse Cresswell, that checks AI use cases for fairness, explainability, privacy and performance before launch and monitors them after deployment. TD Stories says Cresswell leads a team of eight and was TD Inventor of the Year for 2024.
Has TD's responsible AI work won recognition?
TD says its Trustworthy AI approach was named Best Responsible AI Program in the Americas at the DataIQ awards in May 2025, and that it ranked No. 4 for responsible AI in the 2026 Evident AI Index.
Who regulates TD's use of AI in Canada?
OSFI, the Office of the Superintendent of Financial Institutions, whose model-risk Guideline E-23 takes effect on May 1, 2027 and whose July 16, 2026 bulletin covers generative and agentic AI. TD's US bank is covered at /banks/td-bank-us.
- TD strengthens Its Responsible AI Principles to support trust in AI innovation
- Inside the team ensuring artificial intelligence at TD is trustworthy (TD Stories; undated page, retrieved 2026-11-17)
- 3 things to know about breakthrough AI model, TD AI Prism (TD Stories; undated page, retrieved 2026-11-17)
- What's behind AI you can trust? (TD Stories; undated page on the 2026 Evident AI Index, retrieved 2026-11-17)
- Guideline E-23 – Model Risk Management (2027) - Letter
- #4 Maksims Volkovs drives TD Bank's AI transformation
- How TD developed its first agentic AI tool (TD Stories; undated page, retrieved 2026-11-17)
- TD Launches Credit Assist, Turning Days of Commercial Credit Work into Minutes
- TD using AI to speed up mortgage applications and code development
- 5 organizations TD is working with to advance AI innovation (TD Stories; undated page, retrieved 2026-11-17)
- Generative and Agentic Artificial Intelligence: Implications for Technology, Cyber Security, and Operational Resilience
- TD Bank Group 2025 Annual Report: Management's Discussion and Analysis (Form 40-F, Exhibit 99.2)
- TD announces $25 million strategic relationship with Cohere to support AI adoption
- TD Launches Agentic AI to Transform Real Estate Secured Lending from End to End
- TD Bank Group Third Quarter 2026 Report to Shareholders (Form 6-K, Exhibit 99.1)
- How the promise of AI is taking hold at Canada's biggest banks
- TD Bank Group Q4 2025 Earnings Conference Call transcript
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