Banks

HSBC Usa Inc

NY · Large bank ($50B and above)
Total assets of FDIC-insured bank subsidiaries: $165.3B at the end of 2025

Filings on the SEC website · This bank on Bankgraph

In short. In its 2025 annual report, HSBC Usa Inc mentions AI in 6 passages. It lists AI as a risk and explains how AI is controlled.

Compare with peers

In the 2025 annual reportThe yearly report a listed company files with the SEC, called a 10-K. It describes the business, its risks and its results.

Mentions AI
Yes
6 passages
Highest detail levelHow specific a passage is about AI at this bank. General: could be in any bank's report. Names an area: says where AI is used or how it is controlled. Concrete example: names a tool or vendor, gives a number, a date or a result.
Names an area
What it says
Explains how AI is controlled; Sees AI as a risk
Kinds of AI named
Machine learning
How AI is controlled
Model risk management, Policy or framework, Responsible AI, Staff training

AI in its annual reports over time

What this shows
How many passages about AI each annual report contains, 2022 to 2025, by what they say.
What it means
1 passage in 2022, 6 passages in 2025.
How to read it
Each bar is a report year, split by what the passages say. Hover or tap a bar for the count.
Where it comes from
Banks' annual reports (10-K) filed with the SEC, up to 6 Oct 2026. How we did this
Passages about AI in HSBC Usa Inc's annual reports, by report year.
Show as a table
Report yearUsing or planning AIExplains how AI is controlledSees AI as a riskOther mentions
2022
2022
2022
2022
2023
2023
2023
2023
2024
2024
2024
2024
2025
2025
2025
2025

Compared with banks of its size

What this shows
This bank's 2025 annual report next to all 43 banks of its size ($50B and above).
What it means
Its most specific passage is "Names an area"; for banks of its size the typical level is "Names an area".
How to read it
Yes or no for this bank; the share of banks of the same size for comparison.
Where it comes from
Banks' annual reports (10-K) filed with the SEC, up to 6 Oct 2026. How we did this
In the 2025 reportThis bankBanks of its size
Using AI nowNo12 of 43 (28%)
Explains how AI is controlledYes30 of 43 (70%)
Sees AI as a riskYes43 of 43 (100%)
Mentions generative AINo33 of 43 (77%)
Mentions AI agentsNo10 of 43 (23%)

What changed from 2024

0 passages new in the 2025 report, 2 passages from the 2024 report no longer there.

Every passage about AI

What this shows
All 20 passages about AI in this bank's annual reports, quarterly reports and earnings materials since 2023, newest first.
What it means
0 passages say the bank is using AI now.
How to read it
Highlighted words are the terms that matched. Labels show what each passage says. Follow the link to read it in the filing.
Where it comes from
Banks' annual reports (10-K), quarterly reports (10-Q) and earnings materials (8-K) filed with the SEC. How we did this

Annual report, report year 2025 filed 25 Feb 2026

Our cyber hub brings together training, insights, events, and campaigns to provide a one-stop shop on how to combat cyber-crime. We support leaders and model developers with training that helps them understand and apply our principles on the ethical use of big data and artificial intelligence.
Explains how AI is controlledDetail: Names an areaSame as last year
Given the current economic, regulatory, and political environment for large financial institutions such as us, and possible public backlash to bank fees and the emergence of new products, fintech, and non-bank companies, there is increased competitive pressure to provide products and services at current or lower prices. Consequently, our ability to reposition or reprice our products and services from time to time may be limited and could be influenced significantly by the actions of our competitors who may or may not charge similar fees for their products and services. Any changes in the types of products and services that we offer our customers and/or the pricing for those products and services could result in a loss of customers and market share and could materially adversely affect our results of operations. Further, new entrants to the market or new technologies, including artificial intelligence and distributed ledger technology, could require us to spend more to modify or adapt our products to attract and retain customers. Continued technological advances and the growth of e-commerce have made it possible for non-depository institutions to offer products and services that traditionally were banking products, and for financial institutions as well as other companies to provide electronic and Internet-based financial solutions, including electronic payment solutions. We may not respond effectively to these competitive threats from existing and new competitors and may be forced to increase our investment in our businesses to modify, adapt or develop products, services, and technology infrastructure to respond to our customers' needs. Any of these factors may have a material adverse effect on our businesses, prospects, financial condition and results of operations.
Sees AI as a riskDetail: General
A program consisting of policies and procedures that are reasonably designed to achieve compliance with financial crime laws may not always prevent third parties from using us (and our relevant counterparties) as a conduit for money laundering, including illegal cash operations without our knowledge (and that of our relevant counterparties), or committing acts of bribery or violations of economic sanctions through us. In addition, the accessibility and increasing sophistication of artificial intelligence brings financial crime risks. While there is potential for the technology to support financial crime detection, there is also a risk that criminals use artificial intelligence to perpetrate fraud, particularly scams. Developments in payments technologies, including those based on distributed ledger technology, could impact the effectiveness of our financial crime controls. We or our affiliates becoming a party to violations of financial crime laws or regulations, or perceptions of being associated with such activity, could damage our reputation and/or lead to fines, penalties, business restrictions being imposed on us and/or other legal enforcement. Any one of these outcomes could have a material adverse effect on our reputation, business, prospects, financial condition and results of operations.
Sees AI as a riskDetail: GeneralSame as last year
Operational risks are inherent in our businesses and may adversely impact our businesses and reputation. We are exposed to many types of operational risks that are inherent in banking operations, including fraudulent and other criminal activities (both internal and external), breakdowns in processes or procedures and systems failure. For example, fraudsters may target any of our products, services or delivery channels, including lending, internet banking, payments, bank accounts and cards. Further, there is a risk that our operating system controls, business continuity efforts or data security systems could prove to be inadequate. Adoption of emerging technologies, such as artificial intelligence and distributed ledger technology, may also expose us to operational risks, including unanticipated costs, scarcity of specialist expertise, increase in legacy technology burden, reduced resilience or security, or insufficient controls. These operational risks apply equally when we rely on third-party service providers and partners to provide services to us and our customers. These risks may result in financial loss to us (including unanticipated expenses to correct any defect, and litigation expenses), an adverse customer experience, reputational damage and potential regulatory action, which could have a material adverse effect on our businesses, prospects, financial condition and results of operation.
Sees AI as a riskDetail: General
Our operations are subject to risk of cyberattacks. The threat of cyberattacks poses significant risk to financial institutions, including us. Cyberattacks continue to evolve (including due to emerging technologies, such as artificial intelligence) and vulnerabilities in technology systems continue to be exploited by cyber criminals at an increasing rate. Malicious actors, including state and state-sponsored attackers, hackers for hire, organized syndicates, hacktivist groups, and those who might act from within our organization (known as insider threats) have developed a variety of cyberattacks. These attacks include malware (such as ransomware), exploitation of both known and unpublished (zero-day) vulnerabilities in software, data extortion, various types of phishing, distributed denial of service, and other sophisticated methods of compromising critical or sensitive systems and information. Continued geopolitical conflicts and the rising use of artificial intelligence also escalate the overall risk of cyberattacks.
Sees AI as a riskDetail: General
We could incur losses or be required to hold additional capital as a result of model limitations or failure. We use models for a range of purposes in managing our businesses, including regulatory capital calculations, stress testing, credit approvals, financial crime and fraud risk management and financial reporting. We could face adverse consequences as a result of decisions, which may lead to actions by management, based on models that are poorly developed, implemented or used, or as a result of the modeled outcome being misunderstood or the use of such information for purposes for which it was not designed or the inherent limitations arising from the uncertainty in predicting or estimating future outcomes. The adoption of more sophisticated modeling approaches including artificial intelligence and related technology by us and the financial services industry could also lead to increased model related risks. We hold capital for known risks and limitations of our models as appropriate. If additional weakness in a model is discovered or if a model is shown to have failed, we may be required to hold more capital. Risks from use of models could have a material adverse effect on our businesses, financial condition and/or results of operations, minimum capital requirements and reputation.
Sees AI as a riskDetail: Names an areaMachine learningSame as last year

Annual report, report year 2024 filed 19 Feb 2025

Our cyber hub brings together training, insights, events, and campaigns to provide a one-stop shop on how to combat cyber-crime. We support leaders and model developers with training that helps them understand and apply our principles on the ethical use of big data and artificial intelligence.
Explains how AI is controlledDetail: Names an areaSame as last year
we offer our customers and/or the pricing for those products and services could result in a loss of customers and market share and could materially adversely affect our results of operations. Further, new entrants to the market or new technologies, including generative artificial intelligence and distributed ledger technology, could require us to spend more to modify or adapt our products to attract and retain customers. Continued technological advances and the growth of e-commerce have made it possible for non-depository institutions to offer products and services that traditionally were banking products, and for financial institutions as well as other companies to provide electronic and Internet-based financial solutions, including electronic payment solutions. We may not respond effectively to these competitive threats from existing and new competitors and may be forced to increase our investment in our businesses to modify, adapt or develop products, services, and technology infrastructure to respond to our customers' needs. Any of these factors may have a material adverse effect on our businesses, prospects, financial condition and results of operations.
Sees AI as a riskDetail: GeneralGenerative AINew this year
A program consisting of policies and procedures that are reasonably designed to achieve compliance with financial crime laws may not always prevent third parties from using us (and our relevant counterparties) as a conduit for money laundering, including illegal cash operations without our knowledge (and that of our relevant counterparties), or committing acts of bribery or violations of economic sanctions through us. In addition, the accessibility and increasing sophistication of generative artificial intelligence brings financial crime risks. While there is potential for the technology to support financial crime detection, there is also a risk that criminals use generative artificial intelligence to perpetrate fraud, particularly scams. Developments in payments technologies, including those based on distributed ledger technology, could impact the effectiveness of our financial crime controls. We or our affiliates becoming a party to violations of financial crime laws or regulations, or perceptions of being associated with such activity, could damage our reputation and/or lead to fines, penalties, business restrictions being imposed on us and/or other legal enforcement. Any one of these outcomes could have a material adverse effect on our reputation, business, prospects, financial condition and results of operations.
Sees AI as a riskDetail: GeneralGenerative AI
activities (both internal and external), breakdowns in processes or procedures and systems failure. For example, fraudsters may target any of our products, services or delivery channels, including lending, internet banking, payments, bank accounts and cards. Further, there is a risk that our operating system controls, business continuity efforts or data security systems could prove to be inadequate. Adoption of emerging technologies, such as generative artificial intelligence and distributed ledger technology, may also expose us to operational risks, including unanticipated costs, scarcity of specialist expertise, increase in legacy technology burden, reduced resilience or security, or insufficient controls. These operational risks apply equally when we rely on third-party service providers and partners to provide services to us and our customers. These risks may result in financial loss to us (including unanticipated expenses to correct any defect, and litigation expenses), an adverse customer experience, reputational damage and potential regulatory action, which could have a material adverse effect on our businesses, prospects, financial condition and results of operation.
Sees AI as a riskDetail: GeneralGenerative AISame as last year
Our operations are subject to risk of cyberattacks. The threat of cyberattacks poses significant risk to financial institutions, including us. Cyberattacks continue to evolve (including due to emerging technologies, such as generative artificial intelligence) and vulnerabilities in technology systems continue to be exploited by cyber criminals at an increasing rate. Malicious actors, including state-sponsored attackers, hackers for hire, organized syndicates, hacktivist groups, and those who might act from within our organization (known as insider threats) have developed a variety of cyberattacks. These attacks include malware (such as ransomware), exploitation of both known and unpublished (zero-day) vulnerabilities in software, data extortion, various types of phishing, distributed denial of service, and other sophisticated methods of compromising critical or sensitive systems and information. Continued geopolitical conflicts also escalate the overall risk of cyberattacks.
Sees AI as a riskDetail: GeneralGenerative AI
We could incur losses or be required to hold additional capital as a result of model limitations or failure. We use models for a range of purposes in managing our businesses, including regulatory capital calculations, stress testing, credit approvals, financial crime and fraud risk management and financial reporting. We could face adverse consequences as a result of decisions, which may lead to actions by management, based on models that are poorly developed, implemented or used, or as a result of the modeled outcome being misunderstood or the use of such information for purposes for which it was not designed or the inherent limitations arising from the uncertainty inherent in predicting or estimating future outcomes. The adoption of more sophisticated modeling approaches including artificial intelligence and related technology by us and the financial services industry could also lead to increased model related risks. We hold capital for known risks and limitations of our models as appropriate. If additional weakness in a model is discovered or if a model is shown to have failed, we may be required to hold more capital. Risks from use of models could have a material adverse effect on our businesses, financial condition and/or results of operations, minimum capital requirements and reputation.
Sees AI as a riskDetail: Names an areaMachine learningNew this year
•We invested in industry-leading technology and advanced analytic capabilities to improve how we proactively combat financial crime and will continue to monitor regulatory changes that support innovation and effectiveness, including enhancements to our risk management framework to address emerging technology such as artificial intelligence.
Testing or planning AIDetail: Names an areaCompliance and anti-money launderingRisk managementNew this year
•We continue to enhance our model risk framework in response to changes in regulation and external factors. Artificial intelligence and machine learning models remain a key focus. Progress has been made in enhancing governance activity in this area with particular focus on generative artificial intelligence due to the pace of technological change and regulatory and wider interest in adoption and usage.
Explains how AI is controlledDetail: Names an areaMachine learningGenerative AINew this year

Annual report, report year 2023 filed 21 Feb 2024

Our cyber hub brings together training, insights, events, and campaigns to provide a one-stop shop on how to combat cyber-crime. We support leaders and model developers with training that helps them understand and apply our principles on the ethical use of big data and artificial intelligence.
Explains how AI is controlledDetail: Names an areaSame as last year
Given the current economic, regulatory, and political environment for large financial institutions such as us, and possible public backlash to bank fees and the emergence of new products and fintech companies, there is increased competitive pressure to provide products and services at current or lower prices. Consequently, our ability to reposition or reprice our products and services from time to time may be limited and could be influenced significantly by the actions of our competitors who may or may not charge similar fees for their products and services. Any changes in the types of products and services that we offer our customers and/or the pricing for those products and services could result in a loss of customers and market share and could materially adversely affect our results of operations. Further, new entrants to the market or new technologies, including generative artificial intelligence and distributed ledger technology, could require us to spend more to modify or adapt our products to attract and retain customers. Continued technological advances and the growth of e-commerce have made it
Sees AI as a riskDetail: GeneralGenerative AINew this year
A program consisting of policies and procedures that are reasonably designed to achieve compliance with financial crime laws may not always prevent third parties from using us (and our relevant counterparties) as a conduit for money laundering, including illegal cash operations without our knowledge (and that of our relevant counterparties), or committing acts of bribery or violations of economic sanctions through us. In addition, the accessibility and increasing sophistication of generative artificial intelligence brings financial crime risks. While there is potential for the technology to support financial crime detection, there is also a risk that criminals use generative artificial intelligence to perpetrate fraud, particularly scams. Developments in payments technologies, including those based on distributed ledger technology, could impact the effectiveness of our financial crime controls. We or our affiliates becoming a party to violations of financial crime laws or regulations, or accusations of being associated with such activity, could damage our reputation and/or lead to fines, penalties and/or other legal enforcement. Any one of these outcomes could have a material adverse effect on our business, prospects, financial condition and results of operations.
Sees AI as a riskDetail: GeneralGenerative AINew this year
activities (both internal and external), breakdowns in processes or procedures and systems failure. For example, fraudsters may target any of our products, services or delivery channels, including lending, internet banking, payments, bank accounts and cards. Further, there is a risk that our operating system controls, business continuity or data security systems could prove to be inadequate. Adoption of emerging technologies, such as generative artificial intelligence and distributed ledger technology, may also expose us to operational risks, including unanticipated costs, scarcity of specialist expertise, increase in legacy technology burden, reduced resilience, or insufficient controls. These risks apply equally when we rely on third-party service providers and partners to provide services to us and our customers. These operational risks may result in financial loss to us (including unanticipated expenses to correct any defect, and litigation expenses), an adverse customer experience, reputational damage and potential regulatory action, which could have a material adverse effect on our businesses, prospects, financial condition and results of operation.
Sees AI as a riskDetail: GeneralGenerative AINew this year
Our operations are subject to risk of cyberattacks. The threat of cyberattacks poses significant risk to financial institutions, including us. Cyberattacks continue to evolve and vulnerabilities in technology systems continue to be exploited by cyber criminals at an increasing rate. Malicious actors, including state-sponsored attackers, hackers for hire, organized syndicates, hacktivist groups, and those who might act from within our organization (known as insider threat) have developed a variety of cyberattacks. These attacks include malware (such as ransomware), exploitation of both known and unpublished (zero-day) vulnerabilities in software, various types of phishing, distributed denial of service, and other sophisticated methods of compromising critical or sensitive systems and information. An increase in the sophistication of cyberattacks is expected as malicious actors adopt emerging technology, like artificial intelligence, to enhance their methods. Continued geopolitical conflicts also escalate the overall risk of cyberattacks.
Sees AI as a riskDetail: GeneralNew this year

Annual report, report year 2022 filed 21 Feb 2023

Our cyber hub brings together training, insights, events and campaigns to provide a one-stop shop on how to combat cyber-crime. We support leaders and model developers with training that helps them understand and apply our principles on the ethical use of big data and artificial intelligence.
Explains how AI is controlledDetail: Names an area
1 passage in legal noticesThe forward-looking statements notice at the start or end of a filing. It often lists AI among many risks. It is never counted., not counted
•technology changes and developments, including the development, deployment and use of artificial intelligence, including generative artificial intelligence, by HUSI, HSBC Group affiliates, third-party service providers, counterparties, customers, competitors and other market participants, and related operational, cyber, fraud, data, regulatory, compliance, litigation, reputational and competitive risks, as well as such technological developments' impact on how we manage risk;
New this periodNew since the annual report
Quarterly report, page 55Read it in the reportReport an error