How does Northern Trust govern AI responsibly?
Northern Trust says its approach to AI is guided by the NIST AI Risk Management Framework and lists six principles in its 2025 Sustainability Report (August 2026): reliable, secure and resilient systems; accountability, transparency and human oversight; explainability where AI informs decisions; privacy across the AI lifecycle; fairness and mitigation of unintended bias; and security against misuse. Every AI use case goes through a cross-functional review with risk, data privacy and cybersecurity teams, and the bank says it deploys AI only once it meets its standards for security, privacy and resilience, holding third-party providers to the same. The 2025 10-K (February 2026) describes a dedicated AI Risk Program to identify, assess and monitor AI risk, applies the model-risk framework with independent validation to AI and machine-learning models, and names generative AI's 'hallucinatory' output and the EU AI Act and Colorado and California laws as risks. The 2026 proxy says the board oversees AI within strategy, technology and risk, and a Technology and Operations Committee was formed on April 22, 2025. In September 2026 the CFO described a very new governance process that ranks use cases. AI did not meet the bank's 2025 materiality threshold; it has no published standalone AI principles document and no named AI ethics board.[1][2][3][4]
Data as of Oct 16, 2026 · Northern Trust profile updated Oct 13, 2026 · 8 sources cited on this page · 140 banks profiled on this site · Markdown
Which governance bodies and programmes cover AI at Northern Trust?
Each element as the bank describes it in the cited document.
| Element | What it says | Where | Date |
|---|---|---|---|
| AI Risk Program | 'A program to identify, assess, and monitor risks from the use of AI, ensuring AI technologies operate safely, responsibly, and within Northern Trust's risk appetite'; listed with the Model Risk Management Program among operational-risk programmes overseen by the Operational Risk Committee | 2025 10-K | February 24, 2026[2] |
| Model risk framework | Applies to AI, machine learning and other advanced analytics models, with risk-based governance, independent validation and ongoing monitoring | 2025 10-K | February 24, 2026[2] |
| Cross-functional use-case review | All AI use cases are reviewed by teams across risk, data privacy and cybersecurity; AI is deployed only when it meets standards for security, privacy and resilience | 2025 Sustainability Report | August 2026[1] |
| Board oversight | The board oversees AI within strategy, technology and risk and receives periodic AI briefings; Technology and Operations Committee (chair Marcy Klevorn) formed April 22, 2025 | 2026 proxy | March 11, 2026[3] |
| AI monitoring and readiness programme | AI did not meet the criteria for a material sustainability topic in 2025, but the bank 'established a program to support ongoing AI monitoring, governance and readiness' | 2025 Sustainability Report | August 2026[1] |
| Use-case prioritisation | A 'very new' governance process ranks AI projects by productivity and sustainability; the CFO stresses security, reliability and client protection | Barclays conference | September 14, 2026[4] |
| Guardrails group | A central group building AI guardrails while use cases in AML/KYC documents, document digitisation and RFPs go into production | Barclays conference (COO) | September 9, 2025[5] |
What responsible-AI principles has Northern Trust published?
From the 'Responsible AI Governance' panel of the 2025 Sustainability Report, which says the approach is guided by the NIST AI Risk Management Framework.
| Principle | Wording |
|---|---|
| Reliability and security | Designing AI systems to be reliable, secure and resilient, with risks identified and managed[1] |
| Accountability | Maintaining clear accountability, transparency and human oversight[1] |
| Explainability | Supporting explainability where AI informs decisions[1] |
| Privacy | Protecting the privacy of individuals and their data throughout the AI lifecycle[1] |
| Fairness | Promoting fairness and mitigating the risk of unintended bias[1] |
| Security | Strengthening security to guard against misuse and vulnerabilities[1] |
Which AI risks does Northern Trust disclose, and what does it ask of others?
The 10-K says generative AI 'has been known to produce false or "hallucinatory" inferences or output' and can create inaccurate or misleading content and unintended biases that may not be easily detectable, and notes the risk of insufficient rights to data used in or produced by AI tools. It cites the EU AI Act (in force August 1, 2024) and Colorado and California AI laws as regulation the bank monitors.[2]
As an investor, Northern Trust Asset Management says companies should have 'clear AI and cybersecurity risk frameworks/guiding principles, overseen by cross-functional management', with audits and stress testing; its January 2026 stewardship paper on AI informs those expectations.[6]
For family-office clients, Northern Trust's wealth site publishes a template AI responsible-use policy, provided by Andersen Consulting, that sorts AI uses into low, moderate, high and prohibited risk tiers, bans training public models on confidential family data and requires human verification of all AI output.[7]
Not on the record: a standalone AI principles document, a named AI ethics committee, a chief AI officer, or any public count of models under the AI Risk Program. The CFO said in September 2026 that agents must be reliable 'at all times' before they replace staff.[1][4]
What does the Northern Trust profile record on this topic?
The rows below are the entries of the Northern Trust AI strategy profile that bear on this page, as of Oct 13, 2026; the profile holds the full record of 35 dated moves, 17 use cases and 51 sources.
| Use case | What the record shows | Status | Source |
|---|---|---|---|
| ML detection of anomalous and fraudulent transactions | Listed in the 10-K alongside document digitisation and training services teams. | In production | [2] |
| AI Risk Program and model-risk coverage of AI | A dedicated program to identify, assess and monitor AI risk; AI and ML models get risk-based governance, independent validation and ongoing monitoring. | In production | [2] |
| AML/KYC document handling and document digitisation | Use cases the COO described as in production or about to be, built under a central guardrails group; the bank still receives millions of PDF documents from wealth clients and general partners. | Rolling out | [5] |
| AI use-case prioritisation process | A governance process, described as very new in September 2026, ranks AI projects by productivity and sustainability before funding; the CFO stresses security, reliability and client protection. | Rolling out | [4] |
| Board oversight and AI briefings | The board oversees AI inside its strategy, technology and risk work through existing committees, including the Technology and Operations Committee, and receives periodic briefings for directors. | In production | [3] |
| Date | Move | Detail | Source |
|---|---|---|---|
| Sep 14, 2026 | CFO at Barclays: more AI use cases than time and money | A new governance process picks which to fund; adviser tools, next-best-action suggestions, coding support and manual-process automation are in use. | [4] |
| Mar 11, 2026 | 2026 proxy: board oversight of AI | The board oversees AI within strategy, technology and risk and gets periodic AI briefings; the CIO's scorecard credits advancing the enterprise AI programme. | [3] |
| Feb 24, 2026 | 2025 10-K describes AI uses and an AI Risk Program | ML and AI for document digitisation, anomalous and fraudulent transaction detection and staff training; AI models fall under the model-risk framework; generative-AI 'hallucinatory' output named as a risk. | [2] |
| Jan 22, 2026 | 'Scaled AI governance and capabilities' | Listed among 2025 results in the fourth-quarter investor deck, under the productivity pillar. | [8] |
| Sep 9, 2025 | COO at Barclays: $200 million productivity, GitHub saves 31% of developer time | Cherecwich also cites firm-wide Copilot and a guardrails group working on AML/KYC documents, document digitisation and RFPs. | [5] |
Does Northern Trust follow the NIST AI Risk Management Framework?
Yes, by its own account. The 2025 Sustainability Report, published in August 2026, says 'Northern Trust's approach to AI is guided by the National Institute of Standards and Technology (NIST) AI Risk Management Framework' and lists six principles covering reliability, accountability, explainability, privacy, fairness and security.
What is Northern Trust's AI Risk Program?
The 2025 10-K describes it as a programme to identify, assess and monitor risks from the use of AI so that AI 'operate[s] safely, responsibly, and within Northern Trust's risk appetite'. It sits alongside the Model Risk Management Program among operational-risk programmes, and AI models also fall under the model-risk framework with independent validation.
Who oversees AI at Northern Trust's board level?
The 2026 proxy says the full board oversees AI within its strategy, technology and risk oversight and receives periodic AI briefings. The Technology and Operations Committee, chaired by Marcy Klevorn and formed on April 22, 2025, oversees technology, cyber and operating-model strategy.
How does Northern Trust approve new AI tools?
Per the 2025 Sustainability Report, every AI use case goes through a cross-functional review with risk, data privacy and cybersecurity teams, and the bank does not proceed with solutions that fail its standards. The CFO said in September 2026 that a new governance process ranks use cases by productivity and sustainability.
Is AI a material topic for Northern Trust?
Not yet in its sustainability reporting. The 2025 Sustainability Report says AI 'did not meet the criteria for inclusion as a material topic in 2025', while recognising its strategic and regulatory relevance, and that the bank has established a programme for ongoing AI monitoring, governance and readiness.
- Northern Trust 2025 Sustainability Report (Northern Trust and Artificial Intelligence, p. 20)
- Northern Trust Corporation Annual Report on Form 10-K for the fiscal year ended December 31, 2025
- Northern Trust Corporation 2026 Proxy Statement (Schedule 14A)
- Northern Trust at Barclays Conference: Wealth Push Lifts Outlook (transcript)
- Northern Trust at Barclays Conference: Strategic Focus on Growth and Efficiency (transcript)
- Sustainable Investing Trends for 2026
- Artificial Intelligence Responsible Use Policy: template for family offices provided by Andersen Consulting (undated, retrieved 2026-10-15)
- Fourth Quarter 2025 Presentation Materials
When Northern Trust 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