Guidelines for Financial Institutions adopting Enabling Technologies

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) Guidelines for Financial Institutions adopting Enabling Technologies Central Bank of the UAE Securities and Commodities Authority Dubai Financial Services Authority Financial Services Regulatory Authority

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only)
Table of Contents
Introduction ......................................................................................... 3
Objective ............................................................................................ 3
Structure of the Guidelines......................................................................... 4
Scope of Application ............................................................................... 4
Section 1: Definitions .............................................................................. 5
Section 2: Key Principles for adopting Enabling Technologies................................ 10
Section 3: Application Programming Interfaces (APIs) ........................................ 13
Section 4: Cloud Computing ..................................................................... 17
Section 5: Biometrics............................................................................. 25
Section 6: Big Data Analytics and Artificial Intelligence (AI)................................. 27
Section 7: Distributed Ledger Technology (DLT) .............................................. 32
Section 8: Interpretation .......................................................................... 37

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) Introduction The Central Bank of the UAE (CBUAE), together with the Securities and Commodities Authority (SCA), the Dubai Financial Services Authority (DFSA) of the Dubai International Financial Centre and the Financial Services Regulatory Authority (FSRA) of the Abu Dhabi Global Market, collectively referred to as “Supervisory Authorities”, have issued the Guidelines for Financial Institutions adopting Enabling Technologies (“the Guidelines”). The Guidelines are not to be considered as Regulations or Standards being issued by the Supervisory Authorities. The Guidelines are issued pursuant to the powers vested in the respective law of the individual Supervisory Authorities, including:
 The Central Bank of the UAE under the Decretal Federal Law No. (14) of 2018 Regarding the Central Bank & Organisation of Financial Institutions and Activities (“Central Bank Law”);  The Securities and Commodities Authority under Federal Law No. (4) of 2000 concerning the Emirates Securities and Commodities Authority and Market;  The Dubai Financial Services Authority pursuant to the Regulatory Law - DIFC Law No. (1) of 2004 concerning Dubai International Financial Centre; and  The Financial Services Regulatory Authority pursuant to Law No. (4) of 2013 concerning Abu Dhabi Global Market. The increasing adoption of technology-enabled business models presents both opportunities and challenges to those carrying out Innovative Activities. The purpose of the Guidelines is to provide a set of principles when using Enabling Technologies in financial services and accompanying guidance. The key principles are broad enough to cater to the different business models, operating models and financial services offered by existing organizations operating in, and new entrants to, the financial services sector. The key principles are accompanied by more detailed guidance for Institutions to consider when using Enabling Technologies. The Supervisory Authorities may issue further guidance relating to the Guidelines. Objective The objective of the Guidelines is:
 To provide Institutions with best practices on risk management with respect to Enabling Technologies ;  To encourage the safety and soundness of Institutions so that relevant risks arising from innovative business models and services are adequately managed and mitigated;

1 https://www.bis.org/fsi/publ/insights23.pdf https://www.bis.org/bcbs/publ/d431.pdf

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only)  To limit the systemic risks that could arise from the use of innovative technology, thus fostering transparency and financial stability;  To provide guidance on how to manage the risks when adopting Enabling Technologies to deliver more efficient, secure and robust solutions to Customers thereby improving organizational efficiency and financial inclusion; and  To promote the growth and advancement of the UAE financial services sector and encourage adoption of Innovative Activities in the UAE whilst managing risks in a proportionate manner. Structure of the Guidelines The Guidelines are divided into the below sections:
 Section 1: Defines the key terms used throughout the Guidelines;  Section 2: Sets out the key principles relating to the adoption and use of different types of Enabling Technologies; and  Sections 3 to 7: Provides guidance on the application of the key principles covering the use of Application Programming Interface (API), Cloud Computing, Biometrics, Big Data Analytics and Artificial Intelligence (AI) and Distributed Ledger Technology (DLT). Scope of Application The Guidelines are applicable to all Institutions, licensed and supervised by the Supervisory Authorities, who are using or intend to use Enabling Technologies.
Section 2 sets out the key principles which are binding on all Institutions adopting Enabling Technologies
and Sections 3 to 7 cover guidance which Institutions may adopt.
Institutions should apply the Guidelines in a proportionate manner that reflects the size and complexity of the Institution and the nature, scope, complexity and materiality of the Institution’s Innovative Activities. The Guidelines are in addition to any binding regulations, standards, guidance, and other instructions issued by the relevant Supervisory Authorities, which shall take precedence over the Guidelines.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only)
Section 1: Definitions
In the Guidelines, words and expressions have the meaning set out below:
Application Programming
Interface (API)
An API is a set of rules and specifications for software programs to communicate with each other that forms an interface between different programs to facilitate their interaction. There are various types of APIs which include:
 Private APIs: used within an organization to provide interoperability between internal applications in order to help automation and provide flexibility.  Partner APIs: used to integrate software between a company and its partner, often for a very specific purpose like providing a product or service.  Open APIs: Designed to be easily accessible by the wider population where a business relationship may or may not be necessary. This term does not have the same meaning as “Open Banking”. API Consumer Any organization or person who uses the provider’s API to access or send Data. API Lifecycle Refers to the phases of:
a) Conception: the formulation and design of an API; b) Production: the development and testing of an API; c) Publishing: the steps taken to make an API available for use; d) Consumption: the use of the API; and e) Retirement: the withdrawal of the API from use. API Provider An organization that makes APIs available for use by organizations or persons, including the organization itself. Application Application refers to the use of an Enabling Technology in any capacity by an Institution. Artificial Intelligence (AI) AI refers to the theory and development of computer systems able to perform tasks that traditionally use human intelligence. Big Data Analytics Using advanced analytics techniques against a large volume of Data that can be generated by any means and is stored in a digital format.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) Biometrics Automated recognition of individuals based on their biological and behavioral characteristics. It covers a variety of technologies in which unique identifiable attributes of people are used for identification and authentication. These include (but are not limited to) a person’s fingerprint, iris print, hand, face, voice, gait or signature, which can be used to validate the identity of individuals. Biometric attributes are based on an individual’s personal biometric (biological or behavioral) characteristics and typically include the use of one of the following:
 Biophysical Biometric attributes, such as fingerprints, iris patterns, voiceprints, and facial recognition;  Biomechanical Biometric attributes that are the product of unique interactions of an individual’s muscles, skeletal system and nervous system, such as keystroke mechanics; or  Behavioral Biometric attributes that consist of an individual’s various patterns of movement and usage, such as an individual’s email or text message patterns, mobile phone usage, geolocation patterns, file access log etc. Biometrics can be used for the following activities, amongst others:
 Facilitate Customer identification and verification at on-boarding and ongoing Customer authentication;  Support ongoing due diligence and scrutiny of transactions throughout the course of the business relationship;  Provide better and focused Customer services, e.g. identifying regular Customers at the point of entrance and authenticating transactions; and  Aid transaction monitoring for the purposes of detecting and reporting suspicious transactions, as well as, general risk management and anti-fraud efforts. Cloud Computing Use of a network (“cloud”) of hosting processors to increase the scale and flexibility of computing capacity. Such a network could be built by an Institution or made available by a service provider. This model enables on-demand network access to a shared pool of configurable computing resources (for example, networks, servers, storage facilities, applications and services). Customer Customer includes:
 In respect of the CBUAE: A person who is using, or who is or may be contemplating using, any of the services provided by an Institution.  In respect of the SCA: A natural or legal person.  In respect of the DFSA: A Retail Client, Professional Client or Market Counterparty.  In respect of the FSRA: Has the same meaning as “Customer” in Abu Dhabi Global Market.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) Credential Service Provider (CSP) An organization that issues and/or registers authenticators and corresponding electronic credentials (binding the authenticators to the verified identity) to subscribers. CSP’s r maintain subscriber’s identity credential and all associated enrolment Data throughout the credential’s lifecycle and provide information on the credential’s status to verifiers. Data A collection of organised information, facts, concepts, instructions, observations or measurements, in the form of numbers, alphabets, symbols, images or any other form, that are collected, produced, or processed by Institutions. Digital Channels Internet, mobile phones, ATMs (Automated Teller Machines), POS (Point of Sale) terminals, DPAs (Digital Personal Assistants), mobile applications, or other similar means for Institutions to contact other organizations or persons. Distributed Ledger Technology (DLT) DLT refers to a processes and related technologies that enable Nodes in a network (or arrangement) to securely propose, validate, agree and record state changes (or updates) to a synchronised ledger that is distributed across the network’s Nodes. Blockchain is a type of DLT which stores and transmits Data in packages called “blocks” that are connected to each other in a digital ‘chain’. Enabling Technology One of the following types of technologies:
 APIs;
 Cloud Computing;
 Biometrics;
 Big Data Analytics;
 AI; and
 DLT.
Fee
Any fees, charges, penalties and commissions incurred on a product and/or service. Guidelines Refers to these Guidelines. Governing Body Governing Body means an Institution’s Board of Directors, partners, committee of management, or any other form of governing body of a body corporate or partnership. Identity Lifecycle Refers to the phases of:
a) Enrolment: collecting and proofing identity data; b) Issuance: issuing one or more credentials; c) Use: checking identity at the point of transactions; d) Management: maintaining identities and credentials; and e) Retirement: removing the identity record.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) Innovative Activities Technologically enabled provision of financial services which can take various forms and encompass the different sectors of the financial industry (e.g., crowdfunding, payment services). Institution An Institution includes:
 Any Financial Institution licensed and supervised by the CBUAE.
All references to Institutions include any Outsourcing Service Providers acting on behalf of the Institution.  Any entity that has obtained a license or approval from the SCA to engage in a financial activity and / or provide a specific financial service.  Any Authorised Person licensed and supervised by the DFSA. All references to Institutions include any Outsourcing Service Providers acting on behalf of the Authorised Person.  Any Authorised Person or a Recognised Body in Abu Dhabi Global Market. IT Assets Any form of information technology, including software, hardware and Data. Machine Learning Machine Learning is a sub-category of AI that is a method of designing a sequence of actions to solve a problem through learning and experience and with limited or no human intervention. Multi-Factor Authentication Combines use of two (2) or more of the following authentication factors to verify a user’s identity:
 knowledge factor, “something an individual knows”;  possession factor, “something an individual has”;  biometric factor, “something that is a biological and behavioral characteristic of an individual" Nodes Network participants in a DLT network. Outsourcing An agreement with another party either within or outside the UAE, including a party related to the Institution, to perform on a continuing basis an activity which currently is, or could be, undertaken by the Institution itself. The activity should be integral to the provision of a financial service or should be provided to the market by the other party in the name of the Institution. Outsourcing Service Provider A third-party entity that is undertaking an outsourced process, service or activity, or parts thereof, under an outsourcing arrangement. The Institution always remains responsible and accountable for the actions of an Outsourcing Service Provider under an outsourcing arrangement and to ensure that its outsourcing arrangements comply with the principles set out in these Guidelines. Permissioned DLT A distributed ledger which can be updated or validated only by authorised users within set governance rules i.e. special permissions are necessary to read, access or write information on them.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) Permissionless DLT A distributed ledger which can be read or updated by anyone, such as an open-access blockchain used for some cryptocurrencies. Personal Data Personal Data is any information relating to an identified natural person or identifiable natural person. "Identifiable natural person" is defined as a natural person who can be identified, directly or indirectly, in particular by reference to an identification number or to one or more factors specific to their biological, physical, Biometric, physiological, mental, economic, cultural or social identity. Senior Management Senior Management includes:
 In respect of the CBUAE: The executive management of the Institution responsible and accountable to the Board/Governing Body for the sound and prudent day-to-day management of the Institution, generally including, but not limited to, the chief executive officer, chief financial officer, chief risk officer and heads of the compliance and internal audit functions.  In respect of the SCA: A director or group of executives who assume the tasks of planning the daily administrative and supervisory operations of the company's business to achieve its goals and is appointed by the company's Board/Governing Body.  In respect of the DFSA: Includes any senior management and Governing Body within an Authorised Person who take responsibility for that Authorised Person’s arrangements and operations.  In respect of the FSRA: Has the same meaning as “Senior Management” in Abu Dhabi Global Market. Staff One or more employee(s) of the Institution acting in any capacity for or on behalf of the Institution. Subscriber A person whose identity has been verified and bound to authenticators (credentialed) by a Credential Service Provider (CSP). Supervisory Authorities Refers to the following UAE authorities:
 The Central Bank of the UAE (CBUAE);
 The Securities and Commodities Authority (SCA);  The Dubai Financial Services Authority (DFSA) of the Dubai International Financial Centre; and  The Financial Services Regulatory Authority (FSRA) of the Abu Dhabi Global Market. Third Party Any person, group of persons or organization external to, and not a related party to, an Institution.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only)
Section 2: Key Principles for adopting Enabling
Technologies
Application Programming Interfaces (APIs)
2.1 Governance: Institutions should establish an approved and documented governance framework
for effective decision-making and proper management and control of risks arising from the use of APIs.
2.2 Design: Institutions should ensure that APIs (whether designed in-house or by a Third Party) are
designed such that the APIs can flexibly evolve and have robust controls to support cybersecurity, cyber resilience and data protection.
2.3 Management and monitoring: Institutions should establish an approved and documented API
monitoring framework that addresses infrastructure, technology and security related incidents and events in a timely and effective manner.
2.4 Outsourcing: Where an Institution outsources API development to an Outsourcing Service
Provider, the Institution should follow the outsourcing requirements of the relevant Supervisory Authority. Institutions should ensure that the contract governing the arrangement between the Institution and Outsourcing Service Provider contains at a minimum information on the roles and obligations of all parties, liability, dispute management, access to relevant information by the relevant Supervisory Authority, and minimum control measures. Cloud Computing
2.5 Material arrangements: Institutions should assess the materiality and the associated risks of their
Cloud Computing arrangements and address any concerns and expectations that the relevant Supervisory Authority may have prior to implementing any material Cloud Computing arrangement.
2.6 Governance: Institutions should establish an approved and documented governance framework
for effective decision-making and proper management and control of risks arising from the use of Cloud Computing and Outsourcing to Outsourcing Service Providers.
2.7 Auditability: Institutions should ensure the Cloud Computing arrangement is auditable by
maintaining appropriate evidence and records to enable the Institution’s internal control functions, external auditors, regulators and other authorities to conduct their audits and reviews.
2.8 Outsourcing: Institutions should establish an approved and documented governance framework
for outsourcing their Cloud Computing arrangements to appropriately select and monitor vendors as well as mitigate risks arising from Cloud Computing outsourcing arrangements.
2.9 Design: Institutions should implement adequate measures that are commensurate with the
materiality of the arrangement to ensure that Cloud Computing arrangements are resilient, secure, recoverable and meet the capacity and other needs of the Institution.
2.10 Management and monitoring: Institutions should regularly monitor their Cloud Computing
arrangements, to evaluate performance, detect technology and security related incidents, and promptly take any remedial action.
2.11 Data: Institutions should ensure that the use, transmission and storage of data in a Cloud
Computing arrangement complies with applicable laws and regulations and is secured from unauthorised access, use or modification to the extent commensurate with the importance of the data.
2.12 Business continuity: Institutions should put in place a robust and regularly tested business

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) continuity and disaster recovery plan for each material Cloud Computing arrangement and ensure that the plan complies with the relevant Supervisory Authority’s requirements.
2.13 Exit and resolution planning: Institutions should define and maintain specific exit plans for their
material outsourced Cloud Computing arrangements and account for these arrangements when developing resolution plans. Biometrics
2.14 Governance: Institutions should establish an approved and documented governance framework
to control and manage the broad range of risks which may arise from the use of Biometrics.
2.15 Identity proofing and enrolment management: Institutions should establish appropriate identity
verification and proofing mechanisms as part of the Biometric Application’s identity enrolment process.
2.16 Ongoing authentication: Institutions should establish controls and processes to protect the
customer and their credentials against vulnerabilities and unauthorised access, disclosure or use in the authentication process and throughout the Identity Lifecycle.
2.17 Management and monitoring: Institutions should regularly monitor their Biometric Applications
throughout the Identity Lifecycle to evaluate performance, detect security-related events, ensure the adequacy of controls, and promptly take any remedial action.
2.18 Data management: Institutions should ensure the confidentiality, authenticity and integrity of
Data throughout all phases of authentication and whether the Data is in use, storage or transmission. Big Data Analytics and Artificial Intelligence (AI)
2.19 Governance: Institutions should establish an approved and documented governance framework
for effective decision-making and proper management and control of risks arising from the use of Big Data Analytics and AI.
2.20 Accountability: The Governing Body and Senior Management of the Institution should remain
accountable for the outcomes and decisions of their Big Data Analytics and AI Applications including those Applications that make decisions on behalf of the Institutions.
2.21 Design: Institutions should ensure that the models for their material Big Data Analytics and AI
Applications are reliable, transparent and explainable, commensurate with the materiality of those Applications.
2.22 Management and monitoring: Institutions should establish an approved and documented
framework to review the reliability, fairness, accuracy and relevance of the algorithms, models and Data used prior to deployment of a material Big Data Analytics and AI Application and on a periodic basis after deployment, to verify that the models are behaving as designed and intended.
2.23 Ethics: Institutions should ensure that their Big Data Analytics and AI Applications promote fair
treatment, produce objective, consistent, ethical and fair outcomes and are aligned with the Institutions’ ethical standards, values and codes of conduct.
2.24 Customer protection: Institutions should be transparent with Customers about their use of Big
Data Analytics and AI through their conduct and through accurate, understandable and accessible plain language disclosure. Distributed Ledger Technology (DLT)
2.25 Governance: Institutions should establish an approved and documented governance framework
for effective decision-making and proper management and control of the risks arising from the

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) use of DLT.
2.26 Auditability: Institutions should ensure the DLT Application is auditable by maintaining
appropriate evidence and records to enable the Institution’s internal control functions, external auditors, regulators and other authorities to conduct their audits and reviews.
2.27 Design: Institutions should design their DLT Applications to be efficient and effectively secure
IT assets and any customer assets.
2.28 Anonymity and pseudonymity: Institutions developing permissionless DLT Applications should
ensure that users are not anonymous or pseudonymous.
2.29 Management and monitoring: Institutions should ensure their DLT Application are reviewed and
monitored on a periodic basis to evaluate performance, detect technology and security related incidents, ensure the adequacy of controls and promptly take any remedial action.
2.30 Business continuity: Institutions should plan, establish and periodically test arrangements to
maintain the continuity of the service/process performed by the DLT Application in the event of an incident that affects the availability of the Application.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only)
Section 3: Application Programming Interfaces (APIs)
Governance
3.1 Institutions should establish a documented governance framework for effective decision-making
and proper management and control of risks arising from the use of APIs. The governance framework should:
a) Define the roles and responsibilities of the Institution, API Provider and API developer (where different), including the division of duties; b) Establish appropriate policies, procedures, standards and controls to govern the API Lifecycle within the Institution; c) Employ tools and technologies that enable communication, change management and performance monitoring across the API Lifecycle; d) Establish appropriate testing strategies prior to publication and on an ongoing basis for optimal performance of APIs such as:
I) A load testing strategy which can be used to assess how the API performs against service￾level agreements and to determine what response is normal for the API. Target API test case problems that would prevent longer load tests from running correctly should be developed; II) Stress testing of the APIs can be undertaken by simulating a heavy load on the API or by conducting crash point testing to identify the maximum number of users the API can handle; and III) A monitoring framework that ensures critical interfaces and functions are appropriately tested and verified for conformance to expected behavior. e) Establish a framework to assess, monitor, report and mitigate risks associated with the APIs including developing mechanisms to ensure regular testing and implementation of coding controls, production monitoring and support post deployment, process control mapping and development of a risk control matrix; and f) Be approved by the appropriate Governing Body.
3.2 When Outsourcing to an Outsourcing Service Provider, Institutions should ensure access to
information and unrestricted right to monitor, review and audit by the Institution’s internal control functions, and regulators, or persons employed by them, including supervisory reviews by the respective Supervisory Authority; Architecture
3.3 Institutions should ensure that the systems and technology architecture for the APIs are designed
such that the APIs can flexibly evolve. This could be done by making the architecture independent from the applications using the APIs (i.e. such that it is not over tailored to the most common use cases). The evolution of APIs should not hinder existing applications, which should be able to function without interruption.
3.4 Institutions should establish controls so that the architecture supporting the API and the API
itself is secure and protected against misuse or security attacks.
Design
3.5 When determining the design of an API, Institutions may consider the following elements to

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) deliver innovation and flexibility:
a) Accessibility: Ensure all relevant parties can access the API; b) Interoperability: Enable exchange of Data across Institutions without any dependencies on underlying technologies; c) Reuse: Leverage existing standards and taxonomies to avoid duplication of effort; d) Independence: Avoid dependency on any vendors or technologies and retain various options for delivery models and implementation technologies; e) Extensibility: Establish flexibility to extend APIs to new stakeholders and business channels and offer new functionality in existing APIs; f) Stability: Ensure consistency in functionality and accessibility when modifying the API through appropriate governance; g) Transparency: Promote transparency and clarity on the API, including environments supported, changes implemented and standards followed; and h) Loosely coupled: Provide flexibility and minimise impact of changes to operations of other APIs or API applications.
3.6 Institutions should have proper engagement with API Providers before API Providers can
expose any Personal Data through the APIs.
3.7 Institutions should undertake the following steps when designing APIs:
a) Decide on in-house vs outsourced API development; b) Prioritise and sequence APIs to publish; c) Consider guiding principles such as openness, usability and interoperability; d) Ensure adequate security and data protection mechanisms are in place to protect Personal Data; and e) Identify and define requirements and technical guidelines.
3.8 During the development of APIs, Institutions and API Providers should ensure that they:
a) Adopt the appropriate API design model based on the type of the API and the protocol used; b) Develop requirements and technical specification that define the output to be achieved by the APIs and how the APIs should perform their expected functionality from a technical perspective; and c) Document these requirements and technical specifications so that the behavior of the API is well understood and can be measured against expected behavior.
3.9 Institutions should ensure Personal Data being transmitted or stored is encrypted to enable
privacy and integrity of Data. Institutions can consider utilizing secure public/private key based encryption methods and protocols, which should comply with internationally recognised and applicable security standards.
3.10 Institutions should ensure that access management and authentication processes are used so that
only authorised and authenticated individuals and organizations have controlled access to the appropriate API resources.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only)
3.11 Institutions should ensure that authentication mechanisms are implemented effectively and
securely, preventing attackers from compromising authentication tokens or exploiting implementation flaws to assume other user identities temporarily or permanently.
3.12 Institutions should develop an appropriate infrastructure to manage and securely store access
credentials.
3.13 Institutions should use Multi-Factor Authentication when a Customer initially accesses an online
service that uses APIs, to provide secure access to the account.
3.14 Institutions should consider using Multi-Factor Authentication when a customer uses an API to
access, process or transmit Personal Data.
3.15 Institutions should ensure that they create clear access control policies that separate
administrator and regular users and that accurately reflect the hierarchies, groups and roles within the organization. Institutions should review their internal hierarchies, groups and roles to ensure that there are no gaps in the roles that could lead to unauthorized access to the APIs.
3.16 Institutions should ensure that APIs are designed to impose restrictions on the size or number of
resources that can be requested by the user to prevent Denial of Service (DoS) attacks.
3.17 An independent function or external expert with adequate skills and knowledge should conduct
vulnerability assessments and penetration tests on the Institution’s and API Provider’s systems and infrastructure to identify weaknesses or flaws in the security processes at least on an annual basis. Standardisation
3.18 Institutions should consider the adoption of standardised APIs that are issued either by
Supervisory Authorities or the industry. Standardised APIs can, among other matters, include the following:
a) API design standards: Adopting a uniform API design model and language across the relevant financial services industry based on a broad range of design considerations; b) Data standards: Adopting international Data standards that define the semantics and syntax of Data being transmitted using APIs, based on the type of Data being transacted and use case, to promote interoperability; and c) Information security standards: Adopting international information security standards to ensure information is securely transmitted through APIs. Management
3.19 Institutions should consider establishing an API monitoring framework that addresses
infrastructure, technology and security related incidents and events in a timely and effective manner. The monitoring framework should:
a) Define what constitutes an incident and event, such as unusual activity or unauthorised changes; b) Monitor the use of APIs to rapidly and accurately detect incidents and events; c) Report incidents and events to decision makers in a timely manner commensurate with their severity; and d) Remediate the impact of the incidents and events in an effective manner.
3.20 Larger Institutions should consider establishing a security operations center dedicated to

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) monitoring, assessing and defending IT systems and assets such as APIs, web sites, applications, Data servers, networks, hardware and software.
3.21 Institutions should maintain an audit trail that records the appropriate metrics and security￾related behavior of each API and records any breaches of security that occur. The audit trail
should capture the metrics and behavior before and after such breaches to support future detection of breaches of security.
3.22 Institutions should establish incident handling procedures to swiftly detect, review, report and
rectify any incidents. Institutions should only provide necessary details of any incident when reporting incidents to the public to avoid providing attack vectors for bad actors.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only)
Section 4: Cloud Computing
Materiality
4.1 A Cloud Computing arrangement is considered material when a disruption in service or breach
of security or confidentiality of systems and/or Data may have the potential to materially impact:
a) The Institution’s business operations, reputation or profitability; or b) The Institution’s ability to manage risk and comply with applicable laws and regulations; or c) The confidentiality or integrity of an Institution’s or Customer’s Personal Data (i.e. If the arrangement could lead to unauthorized access, disclosure, loss or theft of Personal Data).
4.2 Institutions should conduct an assessment to determine the materiality and the associated risks of
a Cloud Computing arrangement. When conducting such an assessment, Institutions should consider:
a) The criticality and inherent risk profile of the Cloud Computing arrangement i.e. activities that are critical to the business continuity/viability of the Institution and its obligations to Customers; b) The impact and likelihood of a service failure, security breach or other event on an Institution’s business operations or reputation; c) The impact and likelihood of a confidentiality breach, loss or theft of Customer Data or breach of Data integrity of the Institution and its Customers; and d) The cost and other resources to support a Cloud Computing arrangement.
4.3 Institutions should notify the relevant Supervisory Authority of any material Cloud Computing
plans to address any concerns and expectations early in the design process before implementing any material Cloud Computing arrangement. Governance
4.4 Institutions considering the use of Cloud Computing should define a clear strategy and
architectural roadmap which covers the target IT environment, the transition from the current environment to the target and the operating model, including any organizational change or additional skillsets that maybe necessary.
4.5 Institutions should establish an approved and documented governance framework for effective
decision-making and proper management and control of risks arising from the use of Cloud Computing and Outsourcing to Outsourcing Service Providers. The governance framework should:
a) Define the roles and responsibilities for the operation and management of the Cloud Computing arrangement, security controls and risk management controls. Where an Outsourcing Service Provider is involved, the division of roles and responsibility between the Institution and the Outsourcing Service Provider should be clearly defined; b) Define the process to conduct a risk-based analysis to identify and classify the IT Assets involved in or deployed by the Cloud Computing arrangement based on criticality and confidentiality; c) Require the maintenance and updating of the log of IT Assets in the cloud environment including their ownership;

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) d) Establish appropriate policies, procedures and controls to govern the use of Cloud Computing covering risk management, due diligence on the Outsourcing Service Providers and access, confidentiality, integrity and recoverability of IT Assets outsourced; and e) Set out the steps for management and review of the contract between the Institution and the Outsourcing Service Provider, where Cloud Computing services are outsourced.
4.6 Senior Management of the Institution should be responsible for the assessment, understanding
and monitoring of the Institution’s reliance on Outsourcing Service Providers for material Cloud Computing services.
4.7 Institutions should maintain up-to-date and accurate documentation pertaining to the Cloud
Computing arrangement for review, audit, supervision and other purposes, including but not limited to:
a) Rationale and an appropriate strategy for implementing the Cloud Computing arrangement; b) Materiality and risk assessment and conclusion; c) Outsourcing risk assessment, other initial security-related risk assessments and their conclusions (further guidance on assessments provided in subsection “Outsourcing”); d) Due diligence or suitability assessments conducted on the Outsourcing Service Provider and conclusion; e) Description of the Cloud Computing arrangement including but not limited to:
I) Name of Outsourcing Service Provider and any sub-contractors; II) Level of reliance on Outsourcing Service Providers; III) Type of Cloud Computing service models (i.e. Software as a service - SaaS, Infrastructure as a service - IaaS etc.) and deployment models used (i.e. private, public etc.); IV) IT Assets in scope including their criticality and ownership; V) Services/products selected; VI) Parties involved; and VII) Delivery locations. f) Contract and other legal documentation pertaining to the arrangement with the Outsourcing Service Provider (further guidance provided in subsection “Outsourcing”). Outsourcing
4.8 Prior to engaging an Outsourcing Service Provider to provide Cloud Computing services,
Institutions should perform a comprehensive Outsourcing risk assessment covering:
a) The role and materiality of the service to be outsourced in the Institution’s business operations; b) Due diligence on prospective Outsourcing Service Providers (further guidance on the due diligence process provided in clause 4.9); and c) Assessing the benefits of the Outsourcing arrangement against the risks.
4.9 Institutions should verify the maturity, adequacy and appropriateness of the prospective

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) Outsourcing Service Provider and services selected, taking into account the intended usage of the Cloud Computing service. Institutions should consider the following specific factors when conducting due diligence on Outsourcing Service Providers providing Cloud Computing services, including but not limited to:
a) Materiality: The results of the materiality assessment. The depth of the due diligence undertaken and risk mitigating controls established should be commensurate with the materiality of the Cloud Computing arrangement and the level of reliance the Institution places on the provider to maintain effective security controls; b) Due diligence scope: The scope of the due diligence assessment should be appropriate and cover an adequate set of controls and individual assessments of all locations expected to be relevant in the arrangement. In particular, the Institution should consider the track record of the Outsourcing Service Provider in achieving acceptable outcomes in areas such as information security policies and awareness, due diligence and risk assessment of practices related to sub-contracting, system vulnerability assessments, penetration testing, and technology refresh management; c) Data centers: Evaluation of whether the data centers are located in countries that the Institution deems suitable and acceptable to store and process Data (further guidance outlined in the subsection “Design”); d) Controls: Institutions should ensure Outsourcing Service Providers implement strong authentication, access controls, Data encryption and other security and technical controls (further guidance outlined in the subsection “Design” and “Management and monitoring”) to meet the Institutions’ requirements. Controls implemented by Outsourcing Service Providers should be at least as strong as those which the Institutions would have implemented had the operations been performed in-house; e) Security risk assessments: Prior to implementing Cloud Computing services and undertaking an Outsourcing arrangement, Institutions should conduct an initial security and risk assessment of the service to identify any information security, cybersecurity and other IT control weaknesses. The risk assessment will identify security threats including information security threats and operational weaknesses and develop safeguards to mitigate those threats and weaknesses. The factors considered during the risk assessment should include but not be limited to:
I) Nature of the service (including specific underlying arrangements); II) Provider and the location of the service; III) Criticality and confidentiality of the IT Assets involved; IV) Transition process including handover from the Institution and/or other service providers to the potential Outsourcing Service Provider; V) Target operating model; and VI) Adherence to recognised technical security standards. f) Compliance with standards and external assurance: The Outsourcing Service Provider’s adherence to international standards as relevant to the provision of services (for e.g. ISO/EIC etc.). Institutions may take into consideration any external assurance that has already been provided by independent auditors when conducting their own due diligence.
4.10 When conducting risk assessments of Cloud Computing services, Institutions should consider
key risks including but not limited to:
a) Cybersecurity risk;

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) b) Operational risks, specifically information security, Outsourcing and business continuity risk. In particular, Institutions in an outsourced Cloud Computing arrangement should consider the impact of the Outsourcing arrangement on the Institution’s risk profile i.e. the potential heightened operational, legal, compliance, reputational, concentration and other risks associated with the arrangement; c) Reputational risk; and d) Specific risks arising from the design and operating model of the Cloud Computing arrangement.
4.11 Institutions should ensure that the written contract governing the Cloud Computing arrangement
between the Institution and Outsourcing Service Provider covers the following issues including, but not limited to:
a) The roles, relationships, obligations and responsibilities of all contracting parties; b) Region/jurisdiction of the data centers; c) Ownership and control over IT Assets, if the Outsourcing Service Provider is expected to be given some level of control over IT Assets; d) Liability in the event of losses or breaches in security or confidentiality; e) Measures to protect the Institution’s Data and confidential information and limits to disclosure of such information; f) Data recovery and stipulating access to Data used for daily operational purposes as well as for contingency, disaster recovery or backups; g) Advance consent regarding any changes to data center locations; h) Access to information and unrestricted right to monitor, review and audit Cloud Computing arrangements by the Institution’s internal control functions, and regulators, or persons employed by them, including supervisory reviews by the respective Supervisory Authority; i) With respect to Outsourcing Service Providers use of sub-contracting arrangements:
I) Disclosure of all material and service-related sub-contracting arrangements; II) Advance notification of any new sub-contracting arrangements or changes to existing arrangements by the Outsourcing Service Provider; III) Outsourcing Service Provider’s accountability to the Institution for the provision of service and effectiveness of agreed controls; IV) Outsourcing Service Provider’s contractual liability for the performance and risk management of its sub-contractor; and V) The sub-contractor’s full compliance with the obligations existing between the Institution and Outsourcing Service Provider. j) Scenarios or events in which Institutions have the right to terminate the contractual agreement (such as where new or modifications to existing sub-contracting arrangements have an adverse effect on the Institution’s security or risk assessment of the Cloud Computing arrangement); and k) The exit plan and process to be followed in the event of termination of the Cloud Computing arrangement including, but not limited to:

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) I) A reasonable transition period; II) Procedures for returning Data to the Institution; III) Permanent Data deletion by the Outsourcing Service Provider; and IV) Arrangements to transfer the outsourced service to another Outsourcing Service Provider or reincorporate it into the Institution with sufficient hand-over and support from the previous Outsourcing Service Provider.
4.12 Institutions should understand their roles and those of the Outsourcing Service Provider
providing Cloud Computing services. Roles and owners should be defined and agreed upon as
part of the shared responsibility model which should specifically cover roles with respect to
cybersecurity, information security and related controls.
4.13 Where a material Outsourcing arrangement involves the transfer of Data, Institutions should:
a) Classify Data based on criticality and confidentiality; b) Identify potential risks relating to outsourced Data and their impact; and c) Agree on an appropriate level of confidentiality, integrity and availability. Design
4.14 Institutions should ensure the design and architectural aspects of the Cloud Computing services
or arrangement are optimised to cater to the needs of the Institution, adhere to the Institution’s internal policies and procedures and minimise risks.
4.15 Institutions and Outsourcing Service Providers should consider the following principles when
developing the design and architecture of the Cloud Computing arrangement:
a) Availability: To reduce the likelihood of IT Assets becoming unavailable in the event of failure of individual components and improve the ability for users to request and use IT Assets; b) Resilience: To improve resilience through implementation of security controls, use of multiple data centers distributed across multiple locations, use of multiple Outsourcing Service Providers to provide Cloud Computing services and implementation of regular testing and checks to detect security and service issues; c) Recoverability: To allow for swift and effective recovery and restoration of IT Assets to a specified level of service in the event of a compromise of integrity or availability; d) Capacity: To ensure the Cloud Computing arrangement’s capacity is commensurate with the Institution’s needs; and e) Encapsulation: To ensure re-usability of network and system components.
4.16 Institutions should carefully determine and choose the type of cloud(s) deployed based on an
assessment of the business operations performed on the cloud and the risks of each type of cloud.
4.17 Institutions should evaluate and assess the location of data centers while determining the design
of the Cloud Computing arrangement to select data centers appropriate to the Institution’s needs. The assessment should address the location’s:
a) Potential risks, including information security, legal and compliance risks;

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) b) Wider political and security stability; and c) Legislation and legal framework including law enforcement and insolvency law provisions that would apply in the event of an Outsourcing Service Provider’s failure.
4.18 Institutions should implement appropriate and effective network access and security controls
such as firewalls, IPS, advanced threat protection and web proxy so that other on-premise environments are not exposed to unauthorized access from the cloud.
4.19 Institutions should define a standard set of tools and processes to manage containers, images and
release management and ensure consideration of any risks posed by shared virtual environments or Data co-mingling.
4.20 Institutions should ensure that their use, transmission and storage of Data in a Cloud Computing
arrangement complies with all applicable laws and regulations.
4.21 Institutions should implement preventative and detective Data controls to keep Data secure and
prevent Data loss. Institutions should ensure that the Data controls including those outlined in this section cover all Data, whether it is Data in storage, Data in transmission (i.e. Data that is actively moving from one location to another) or Data in use.
4.22 Institutions should ensure Data processed or stored through the Cloud Computing arrangement is
recoverable within a pre-defined timeframe and appropriate and secure backups of Data are maintained.
4.23 Where the Cloud Computing arrangement is using a multi-tenancy environment or Data
commingling arrangement, Institutions should ensure its Data and information is segregated and the Outsourcing Service Provider is able to protect the confidentiality and integrity of the Data and information.
4.24 Institutions should introduce controls to prevent unauthorised access to Data and permit access
to IT Assets only when appropriate.
4.25 Institutions should establish security controls to protect against attacks (e.g. network intrusion
attempts, DoS attacks) including cloud specific attacks.
4.26 Institutions should introduce cryptographic key management to control access to, segregate and
secure Customer’s Data.
4.27 Institutions should utilise encryption or tokenisation to protect confidentiality of Personal Data,
such as authentication credentials and emails etc., being processed, or in transit including Data in Data back-ups.
4.28 Institutions should introduce user identity and access management and authentication (including
Multi-Factor Authentication) to provide controlled access to information systems allowing Staff and Outsourcing Service Providers to perform their business activities, while protecting the Data and systems from unauthorised access.
4.29 Institutions should ensure that user access and activities are logged and reviewed on an as is
needed basis.
4.30 Institutions should develop controls to ensure confidentiality and integrity of source codes and
prevent alteration of source codes and system configurations (particularly when the Institution uses models such as DevOps).
4.31 Institutions should conduct vulnerability assessments and penetration tests specific to the Cloud
Computing arrangement to identify weaknesses or flaws in the security processes. Management and monitoring

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only)
4.32 Institutions should establish change management processes to ensure any changes in the Cloud
Computing arrangement by the Institution or the Outsourcing Service Provider are appropriately governed and implemented.
4.33 Institutions should ensure that they define the conditions and scenarios in which automated
testing and releases can take place for changes to their Cloud Computing arrangements. Institutions should ensure that there is a full audit trail, record of the changes and evidence of pre-approval.
4.34 Institutions should develop a mechanism by which they are notified of material changes to the
Cloud Computing arrangement in a timely manner.
4.35 Institutions should develop a configuration management process which includes regular
monitoring to detect unauthorised changes to the cloud environment and ensure such changes can be appropriately remediated.
4.36 Institutions should ensure the Cloud Computing arrangement have the capacity to run the
Institution’s workloads. Institutions should regularly monitor utilisation and proactively plan for upgrades or enhancements based on anticipated spikes in workloads or strategic business initiatives.
4.37 Institutions should establish a monitoring framework to define, monitor, report and remediate
key infrastructure, technology and security related incidents and events in the cloud environment in a timely and effective manner to minimise detriment. The framework should:
a) Cover incidents and events that may impact the stability or availability of the Institution’s applications, networks and systems or the confidentiality or integrity of cloud environments; b) Be centralised to promote clarity of process and enable consolidation and analysis of threat intelligence, incident and event related Data; c) Manage incidents and events according to their frequency, criticality and assigned ownership; d) Identification, monitor and manage systemic issues; e) Monitor and identify vulnerabilities, incidents and events on an on-going basis by:
I) Defining a standard set of health and performance metrics; II) Utilising analytics and Data from previous security incidents and events to enable retrospective detection; f) Categorise and record Data associated with incidents and events; g) Report and escalate incidents and events to relevant stakeholders for notification or action; and h) Ensure incidents and events are properly reviewed and identified gaps are remediated to prevent a reoccurrence.
4.38 Institutions should be able to swiftly and safely:
a) Detect vulnerabilities in the software used in the cloud environment; and b) Deploy security and operating system patches.
4.39 After implementation of the Cloud Computing arrangement, Institutions should re-assess the risk
associated with the Cloud Computing arrangement when there is a material change to existing

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) arrangements and on a regular basis through:
a) On-going Outsourcing risk assessments to assess adequacy of controls in managing the risks arising from the Outsourcing arrangement; and b) On-going security and risk assessments to assess the adequacy of the security and risk controls in managing the risks arising from Cloud Computing. This should include conducting vulnerability assessments and penetration tests specific to the Cloud Computing arrangement on at least an annual basis.
4.40 Institutions should establish risk mitigation controls to address any shortcomings of the Cloud
Computing arrangement. The degree of risk should inform the stringency of controls and mitigation procedures implemented. Business continuity
4.41 Institutions’ business continuity management function and crisis management team should
develop and implement a business continuity and disaster recovery plan for material Cloud Computing arrangements. If Cloud Computing arrangements are outsourced, the Outsourcing Service Provider should have a business continuity and disaster recovery plan in place that is acceptable to the Institution.
4.42 Institutions should define key risk indicators, performance metrics and adverse conditions that
can trigger the business continuity plan for the Cloud Computing arrangement during its on￾going monitoring and oversight of any services provided by the Outsourcing Service Provider.
4.43 AError! Reference source not found.s part of its own business continuity and disaster
recovery planning for Cloud Computing services, Institutions should tailor the plans to:
a) Account for any dependency on one Outsourcing Service Provider; b) Define the division of roles and responsibilities; c) Define recovery objectives; d) Identify alternative solutions/develop transition plans; and e) Test their business continuity plans and disaster recovery plans for their Cloud Computing arrangement (jointly with the Outsourcing Service Provider if the Cloud Computing arrangement is outsourced) on at least an annual basis. Exit and resolution planning
4.44 Institutions should consider the possibility of a stressed exit wherein an event of disruption
cannot be managed through business continuity measures.
4.45 Institutions should define and maintain specific exit plans for their outsourced Cloud Computing
arrangements, taking into account developments (such as new technology) that may change the feasibility of an exit in stressed and non-stressed scenarios.
4.46 Institutions should account for outsourced Cloud Computing arrangements when developing
resolution plans or strategies to identify and address any impediments to its resolvability and to prepare for its possible resolution.
4.47 Institutions should establish procedures for Data recovery by the Institution and permanent Data
deletion by the Outsourcing Service Provider in the event of a termination of services or exit.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only)
Section 5: Biometrics
Governance
5.1 Institutions should establish an approved and documented governance framework to govern the
reliability of the technology used and ensure that adequate policies, procedures and controls are implemented in order to manage risk associated with the Biometric Application.
5.2 When Outsourcing to an Outsourcing Service Provider, Institutions should ensure access to
information and unrestricted right to monitor, review and audit by the Institution’s internal control functions, and regulators, or persons employed by them, including supervisory reviews by the respective Supervisory Authority. Identity proofing and enrolment management
5.3 Institutions using Biometric Applications should ensure effective proofing of identities,
including validation and verification of identities. Validation involves determining that the documentary evidence for the Biometric identity is genuine, reliable and independent. Verification involves confirming the validated identity relates to the individual being proofed.
5.4 Institutions should obtain and store evidence of any digital identity verification (e.g., via chip or
wireless technologies) performed by integrated scanners, sensors and other devices. Ongoing authentication and identity life cycle management
5.5 Institutions should establish controls and processes to protect the customer and their credentials
against vulnerabilities and unauthorized access, disclosure or use in the authentication process and throughout the Identity Lifecycle.
5.6 Institutions deploying Multi-Factor Authentication at login that includes a Biometric factor
should consider employing phishing-resistant authenticators where at least one factor relies on public key encryption to secure the Customer authentication process.
5.7 Institutions should implement risk-based or adaptive authentication measures that present
Customers with authentication options commensurate with the risk level of the transaction and sensitivity of the information.
5.8 Institutions should implement Multi-Factor Authentication using a Biometric factor, where
possible, to authorize high risk activities and protect the integrity of Customer account Data and transaction details. High-risk activities include changes to Personal Data (e.g. Customer office and home address, email and telephone contact details), registration of Third-Party payee details, high value funds transfers and revision of funds transfer limits.
5.9 For Biometric authentication of transactions, Institutions should adopt security measures to
ensure the confidentiality, authenticity and integrity of authentication codes, Personal Data and transaction specific information. Management and monitoring
5.10 Institutions should periodically monitor their Biometric Applications throughout the Identity
Lifecycle to assess performance, detect security-related events, evaluate the adequacy of controls, and take any remedial action.
5.11 Institutions should ensure that all communications with individuals involving Biometric Data
across the Identity Lifecycle occur over a mutually authenticated and protected channel. Institutions should ensure the employment of physical tamper detection and resistance features appropriate for the environment in which the identity-proofing session occurs.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only)
5.12 Throughout the Identity Lifecycle, Institutions should introduce processes and controls to
safeguard against Data tampering, cyberattacks, security breaches and other fraudulent activities which may lead to identity theft, compromise or misuse of Data and errors.
5.13 Institutions should monitor and evaluate all the processes involved in the Identity Lifecycle
including identity proofing, authentication etc. to ensure that they are secure and efficient.
5.14 As a CSP may be an independent Third Party or may issue credentials for its own use (e.g., large
Institution), Institutions should ensure that they perform the requisite due diligence checks and protocols on the CSP on a regular basis.
5.15 Institutions should monitor the performance of the Biometrics Application for inherent risks
such as false acceptance rates and false rejection rates. Poorly executed algorithms may result in higher false acceptance rates and these inherent risks should be calibrated to be commensurate with the risk associated with the Biometric Application. Data management
5.16 Institutions should ensure the confidentiality, authenticity and integrity of Data across all phases
of authentication, whether the Data is in use, storage or transmission.
5.17 Institutions should maintain a clear trail and record of the Biometric Application’s Data obtained
from the CSPs.
5.18 Institutions should consider the principles of portability and interoperability when planning and
implementing the systems and databases for the Biometric Application.
5.19 Where Biometric Applications are used for Biometric authentication, Institutions should ensure
Biometric Data and authentication credentials, whether in use, storage or transmission, are encrypted. Outsourcing
5.20 If Biometric activities are outsourced, Institutions should ensure that:
a) They obtain the necessary Data concerning Biometric identification/verification from the Outsourcing Service Providers and take adequate steps to satisfy themselves that copies of identification Data and other relevant documentation will be made available from the Outsourcing Service Providers upon request and without delay; and b) The Outsourcing Service Provider complies with the Institution’s Customer due diligence and record-keeping requirements.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only)
Section 6: Big Data Analytics and Artificial Intelligence
(AI)
Materiality
6.1 Institutions should assess their Big Data Analytics and AI Applications to determine the
materiality and associated risks of each Application.
6.2 When conducting a materiality assessment of a Big Data Analytics and AI Application,
Institutions should consider:
a) The purpose of the Big Data Analytics and AI Application (i.e. use case) and its role in the Institution’s decision-making process; b) The criticality and inherent risk profile of the activities (i.e. are they activities that are critical to the business continuity/viability of the Institution and its obligations to Customers); and c) The likelihood that the activity may be disrupted and the impact of a disruption. Governance
6.3 Institutions should establish an approved and documented governance framework for effective
decision-making and proper management and control of risks arising from the use of Big Data Analytics and AI. The governance framework should:
a) Establish a mechanism to ensure prior to adoption of Big Data Analytics and AI for a particular Application, Institutions assess whether the Application is suitable for Big Data Analytics and AI implementation. Institutions should define specific parameters and criteria to enable the Institution in its decision-making; b) Establish appropriate policies, procedures and controls to govern the design, development, monitoring, review and use of Big Data Analytics and AI Applications within the Institution; c) Ensure proper validation of Big Data Analytics and AI Applications prior to their launch, and thereafter implement on-going training, calibration and review to ensure the reliability, fairness, accuracy and relevance of the algorithms, models, Data used and the results; d) Maintain a transparent, enterprise-wide record of Big Data Analytics and AI Applications and their underlying mechanics; e) Establish processes to assess, monitor, report and mitigate risks associated with the Big Data Analytics and AI Application; f) Ensure material decisions regarding Big Data Analytics and AI Applications and their underlying models and Data are documented and sufficiently justified; and g) Cover every stage of the model lifecycle including design, development, deployment, review, update and discontinuation.
6.4 The Governing Body and Senior Management of the Institution should be accountable for the
outcomes and decisions of their Big Data Analytics and AI Applications, including those Applications that make decisions on behalf of the Institution. They should:
a) Ensure that all Staff have been assigned appropriate accountability for their involvement with Big Data Analytics and AI Applications and understand what they should to do meet this accountability; and

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) b) Ensure technical specialists with appropriate technology skillsets (e.g. Big Data analysts, Artificial Intelligence engineers and specialists) and Application specific skillsets (e.g. credit risk modelling specialists if the Application is a credit scoring model) form part of the team actively involved in developing and implementing Big Data Analytics and AI Applications.
6.5 When Outsourcing to an Outsourcing Service Provider, Institutions should ensure access to
information and unrestricted right to monitor, review and audit by the Institution’s internal control functions, and regulators, or persons employed by them, including supervisory reviews by the respective Supervisory Authority.
6.6 Big Data Analytics and AI Applications, including when the model is developed by an
Outsourcing Service Provider, should be auditable and, accordingly, Institutions, where relevant and taking into account the type of application used, should maintain on-going and up-to-date information through:
a) Establishing audit logs and maintaining traceability of decisions and outcomes of the Big Data Analytics and AI Application; b) Developing and maintaining design documentation (further guidance provided in clause 6.12); c) Maintaining records of the various versions of the model including its code (further guidance provided in clause 6.17); d) Archiving original Datasets used to develop, re-train or calibrate models; e) Maintaining replication protocols and instructions; f) Tracking outcomes and performance of the Big Data Analytics and AI Application; and g) Retaining above information for a minimum period of five (5) years, or as otherwise prescribed by applicable laws and regulations. Design This sub-section applies only to material Big Data Analytics and AI Applications.
6.7 Institutions should ensure that the models for their material Big Data Analytics and AI
Applications are reliable, transparent and explainable, commensurate with the materiality of those Applications. Accordingly, Institutions, where relevant, should consider:
a) Reliability: Implementing measures to ensure material Big Data Analytics and AI Applications are reliable and accurate, behave predictably, and operate within the boundaries of applicable rules and regulations; b) Transparency: Institutions should be transparent in how they use Big Data Analytics and AI in their business processes, and (where reasonably appropriate) how the Big Data Analytics and AI Applications function; and c) Explainability: Implementing measures to ensure the technical processes and decisions of a Big Data Analytics and AI model can be easily interpreted and explained to avoid the threat of “black-box” models. The level of explainability should be appropriate and commensurate with the purpose and materiality of the Big Data Analytics and AI Application (e.g. where the model results have significant implications on decision making).
6.8 Institutions should adopt an effective Data governance framework to ensure that Data used by
the material Big Data Analytics and AI model is accurate, complete, consistent, secure and provided in a timely manner for the Big Data Analytics and AI Application to function

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) appropriately.
6.9 Institutions should make regular efforts to ensure data used to train the material Big Data
Analytics and AI model is representative (i.e. how relevant the Data and inferences drawn from the Data are to the Big Data Analytics and AI Application) and produces predictable, reliable and optimised outcomes.
6.10 Institutions should be able to promptly suspend material Big Data Analytics and AI Applications
upon the Institution’s discretion such as in the event of a high risk cyber threat, information security breach or malfunctioning of the model.
6.11 Institutions should, where relevant, conduct rigorous, independent validation and testing of
material trained Big Data Analytics and AI models to ensure the accuracy, appropriateness and reliability of the models prior to deployment. Institutions should ensure the model is reviewed to identify any unintuitive or false causal relationships. The validation may be carried out by an independent function within the Institution or by an external organization.
6.12 Institutions should maintain documentation outlining the design of the material Big Data
Analytics and AI model including but not limited to, where applicable:
a) The input Data source and Data description (types and use of Data); b) The Data quality checks and Data transformations conducted; c) Reasons and justifications for specific model design and development choices; d) Methodology or numerical analyses and calculations conducted; e) Results and expected outcomes; f) Quantitative evaluation and testing metrics used to determine soundness of the model and its results; g) Model usage and implementation; h) Form and frequency of model validation, monitoring and review; and i) Assumptions or limitations of the model with justifications.
6.13 Institutions should introduce controls to ensure confidentiality and integrity of the codes used in
the material Big Data Analytics and AI Application so that the code is only accessed and altered by authorized persons.
6.14 Institutions should identify and monitor the unique risks arising from use of the material Big
Data Analytics and AI Application and establish appropriate controls to mitigate those risks. Management and monitoring This sub-section applies only to material Big Data Analytics and AI Applications.
6.15 Institutions should establish an approved and documented framework to review the reliability,
fairness, accuracy and relevance of the algorithms, models and Data used prior to deployment of a material Big Data Analytics and AI Application and on a periodic basis after deployment, to verify that the models are behaving as designed and intended. The framework should cover, where relevant:
a) The various types and frequencies of reviews including continuous monitoring, re-training, calibration and validation;

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) b) Scenarios and criteria that would trigger a re-training, calibration, re-development or discontinuation of the model such as a significant change in input Data or external/economic changes; c) Review of material Big Data Analytics and AI model outcomes for fairness or un-intentional bias (e.g. through monitoring and analysis of false positive and/or false negative rates); and d) Review of continuity or contingency measures such as human intervention or the use of conventional processes (i.e. that do not use Big Data Analytics and AI).
6.16 When the use of a material Big Data Analytics and AI model results in a technical or model￾related error or failure, Institutions should:
a) Be able to swiftly detect the error; b) Establish a process to review the error and rectify it in a timely manner, which may include notifying another function; and c) Report the error to relevant stakeholders, if material.
6.17 Institutions should establish a robust system for versioning and maintain record of each version
of the material Big Data Analytics and AI model including but not limited to, where applicable:
a) New Data used; b) Revisions to the documentation; c) Revisions to the algorithm; d) Change in the way variables are picked and used in the model or, where possible, the names of variables; and e) The expected outcome of the newly calibrated, re-trained or re-developed model. Ethics
6.18 Institutions should ensure that their Big Data Analytics and AI Applications promote fair
treatment produce objective, consistent, ethical and fair outcomes and are aligned with the Institutions’ ethical standards, value and codes of conduct. Accordingly, they should:
a) Comply with laws against discrimination and other applicable laws; b) Be produced using representative inputs and Data which have been tested for selection bias (further guidance provided in clause 6.9); c) Consider whether a human-in-the-loop mechanism is needed to detect and mitigate biases; d) Retain the possibility of manual intervention to mitigate or reverse irresponsible and erroneous decisions; e) Retain the possibility of modification by the Institution; and f) Be explainable (further guidance in clause 6.7c).
6.19 Institutions should consider the fairness of a Big Data Analytics and AI model through
understanding the biases and noise affecting Big Data Analytics and AI decisions and the trade￾offs involved in different conceptions of fairness. Institutions should define what it means for a Big Data Analytics and AI model to be fair.
6.20 Institutions should consider and assess the impact that Big Data Analytics and AI models may

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) have on individuals or groups of individuals to ensure that such individuals or groups are not systematically disadvantaged unless the decisions suggested by the models have a clearly documented justification. Institutions should take steps to minimize unintentional or undeclared bias.
6.21 Institutions should consider restricting the Big Data Analytics and AI model’s use to
environments in which they are likely to be reliable or unbiased (i.e. a chatbot could be designed to avoid collecting Personal Data or ignore certain words to reduce the risk of manipulation), where possible. Customer protection
6.22 Institutions should be transparent with Customers about their use of Big Data Analytics and AI
through their conduct and through accurate, understandable and accessible plain language disclosure. Institutions should:
a) Ensure Customers are informed of Products and/or Services that utilise Big Data Analytics and AI and the associated risks and limitations of the technology, prior to providing the service or each time Customers interact with the service (e.g. in the case of a Customer￾facing service); b) Explain how to use the Big Data Analytics and AI Application to Customers and ensure Customers have easy access to the instructions at all times; and c) Provide explanations of the types of Data, types of variables and decision-making process used by Big Data Analytics and AI Applications upon Customers’ request.
6.23 Institutions should obtain each Customer’s acceptance of the risks associated with the use of Big
Data Analytics and AI, prior to providing the service.
6.24 Institutions should put in place a mechanism for Customers to raise inquiries about Big Data
Analytics and AI Applications and request reviews of decisions made by Big Data Analytics and AI Applications.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only)
Section 7: Distributed Ledger Technology (DLT)
Governance
7.1 Institutions should establish an approved and documented governance framework for effective
decision-making and proper management and control of risks arising from the use of DLT. The governance framework should include the following, as may be relevant depending on the type of DLT:
a) Cover the following elements integral to the functioning of a DLT Application:
I) Ownership model of the DLT platform and the Nodes running on it; II) Model used to operate and manage the distributed ledger (i.e. a consortium, a single Institution); III) Rules to govern the ledger including participant and validator rules and restrictions; IV) Approval processes and procedures to grant access to create, read, update or deactivate Data stored on the distributed ledger; V) Managing public and private keys; VI) Consensus protocol; and VII) Off-chain procedures (if any) including parameters for the validity of an off-chain activity and any standards or requirements for off-chains systems are defined and complied with. b) Define the roles and responsibilities of the key groups involved with respect to the design, development and operation of the distributed ledger. Key groups may include:
I) Core group who will design, govern and operate the distributed ledger; II) Qualified users of the distributed ledger, such as other Institutions and miners; III) Participants of the distributed ledger, such as owners of cryptocurrency etc.; and IV) Third Parties including Outsourcing Service Providers such as custodians or software developers involved in delivering the service.
7.2 Reviews of the DLT Application should be conducted with oversight from Senior Management,
prior to launch and thereafter on an on-going basis to ensure its reliability and security.
7.3 Institutions should establish clear and unambiguous governing rules for participants of the
distributed ledger for onboarding, on-going operations and dispute resolution.
7.4 When Outsourcing to an Outsourcing Service Provider, Institutions should ensure access to
information and unrestricted right to monitor, review and audit by the Institution’s internal control functions, and regulators, or persons employed by them, including supervisory reviews by the respective Supervisory Authority.
7.5 Institutions should ensure their DLT Applications maintain appropriate evidence and records to
enable the Institution’s internal control functions, external auditors, regulators and other authorities to conduct their audits and reviews. Accordingly, Institutions should:
a) Record and store the additional evidence and information to provide auditors with a complete representation of processes, internal controls, financial statements, etc., and for

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) proper accounting treatment of the transaction; b) Ensure the distributed log of records of the DLT Application are fully available and accessible to the relevant parties to audit and review; c) In the event that the DLT is in the form of a blockchain, ensure off-chain activities, rules and protocols associated with and any link to on-chain activities are recorded and stored; and d) Ensure the DLT code and subsequent updates are recorded and stored. Design
7.6 Institutions should design their DLT Applications to be efficient and effectively secure IT assets
and any customer assets. Institutions should, where possible, ensure the design and architectural aspects of the DLT are optimised to cater to the specific use of the technology and the needs of the Institution.
7.7 Institutions should consider the following principles when developing the design of a DLT
Application that manage financial assets:
a) Flexibility: To define the DLT Application narrowly, which is easier to optimise, or flexibly, which may incur higher costs (more difficult protocol, less predictability, larger target for attacks, etc.); b) Traceability: To ensure the distributed log of records and any off-chain records are traceable and anonymity is avoided; c) Capacity: To ensure the DLT Application’s computing and Data capacity is commensurate with the Institution’s needs and is scalable depending on the intended use; d) Security: To ensure the activities and records on the distributed ledger are secure; e) Confidentiality: To maintain confidentiality of Data and control the number of Nodes that increase the risk of confidentiality; and f) Resilience: To ensure the system is resistant to Data loss, loss of integrity lack of availability, or manipulation. In general, a higher number of Nodes improves the resilience of the system.
7.8 Institutions should define rules relating to Data and technological architecture of the DLT
Application during the design phase to ensure that internal control functions, external auditors and Supervisory Authorities can effectively access the application (when applicable) and monitor compliance.
7.9 Institutions should appropriately determine the type of access, permissioned or permissionless,
granted to participants based on an assessment of the operations performed, level of security and Data stored using DLT. For instance, Applications involving any of the following elements should use permissioned systems:
a) Customer assets, funds or other forms of ownership, rights or interests such as contracts; b) Personal Data; or c) Requirement for a controlled set of participants or restricted access.
7.10 Institutions should establish controls to ensure the integrity and security of the network and DLT
Application. These controls may include, but are not limited to:
a) Implementing processes to limit Node processing ability and make it difficult for a Node to

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) process a large number of transactions; b) Designing the Application to allow blocking of IPs/Nodes that generate too many new transactions; and c) Institutions developing permissioned DLT Applications should ensure individual Nodes comply with security standards and requirements to guard against attacks.
7.11 Institutions should establish appropriate consensus protocols or platforms to accept and validate
new records. The consensus methodology should be based on the Institution’s requirements with respect to performance, scalability, consistency, Data capacity, governance, security and failure redundancy.
7.12 Institutions should avoid storing clear-text Personal Data on a blockchain and use sidechains or
other private storage options.
7.13 Institutions should introduce cryptographic key management to control access to confidential
information and Personal Data. Institutions should institute controls with respect to key generation and management and ensure an appropriate and secure storage and transmission mechanism for private keys.
7.14 With respect to issuance of keys to Staff and other personnel:
a) Institutions should issue individual keys to Staff and other persons working on behalf of the Institution for audit, supervision and review purposes; b) Institutions should embed internal checks and verifications on the transactions and activities executed through the distributed ledger such as Staff sign-off on requests or transactions; and c) Institutions should ensure they can internally identify the Staff signing-off messages or requests on the distributed ledger.
7.15 Institutions should, if possible, ensure the design of the DLT Application allows for change i.e.
add new functionality or remove unwanted or incorrectly functioning features within the DLT Application.
7.16 Institutions should, if possible, should introduce mechanisms to manage forks (i.e. conflict
where there are incompatible versions of the DLT broadcast at about the same time) and ensure that the situation is resolved quickly and the integrity of the DLT is maintained.
7.17 Institutions should, if possible, introduce user access management and authentication to provide
controlled access to the DLT Application. Staff and Outsourcing Service Providers should be provided access to only those parts of the system they need to perform their business activities. Institutions may consider if, in certain cases it might be feasible to accept transactions only from select, authorised IP addresses.
7.18 Institutions should develop controls to ensure confidentiality and integrity of codes and prevent
alteration of codes.
Anonymity and pseudonymity
7.19 Institutions developing permissionless DLT Applications should ensure that users are not
anonymous or pseudonymous. Allowing anonymity and pseudonymity can facilitate criminal purposes like tax evasion, bribery, money laundering or terrorism financing.
7.20 Further, Institutions should consider using mechanisms such as chain analytics to follow and
monitor transactions on pseudonymous blockchains.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) Management and monitoring
7.21 Institutions should ensure their DLT Applications are reviewed and monitored on a periodic
basis to evaluate performance, detect technology and security related incidents, ensure the adequacy of controls and promptly take any remedial action.
7.22 Institutions should ensure the code is reviewed and validated on an ongoing basis at regular,
predetermined intervals to identify any weaknesses. Institutions should ensure the code is performing its expected functions and is secure.
7.23 Institutions should conduct vulnerability assessments and penetration tests specific to the DLT
Application to identify weaknesses or flaws in the security processes.
7.24 Institutions should manage and monitor information integrity, privacy and confidentiality in the
implementation of DLT throughout its lifecycle (e.g. by conducting privacy threshold analysis, privacy impact assessment, and Data protection impact assessment). Institutions may adopt processes such as sharding and pruning during the design of the Application to further manage privacy and confidentiality mechanisms.
7.25 Institutions should further ensure control and monitoring of the DLT through an even stringent
focus on encryption than traditional controls with a particular focus on key management. A related control to consider is encrypting the ledger with more than one key and applying on￾chain encryption.
7.26 Institutions should review the distributed log of records and transactions within the ledger to
identify suspicious patterns and connections to monitor any anomalous behavior using analytics and testing.
7.27 Institutions should adopt operations security controls including standard infrastructure controls
such as virus checking schedules, 0-day exploit remediation, maintenance schedules, capacity and backup management.
7.28 Institutions should adopt security incident management controls that describe the processes
around reporting, escalation and response to any breaches. Institutions should monitor if one of the Nodes increases processing power and is executing a significantly higher number of transactions.
7.29 Institutions should ensure human resources put security controls in place to monitor access to the
DLT Application and system. Institutions should ensure these controls are updated when Staff leave or change roles.
7.30 Institutions should maintain and monitor physical and environmental security through use of
hardware security modules, physical security measures such as CCTVs, physical barriers, traditional key security and access controls. Data standardisation and interoperability
7.31 Institutions should not maintain Personal Data on the ledger and such Data should be maintained
off-chain.
7.32 Data retention will need to be factored in the underlying design of the network for Nodes to
purge ledger information after certain defined time periods. Where Data retention rules apply to individual Data sets, destruction of keys used to encrypt the on-chain Data should be implemented.
7.33 Depending on the Application, Institutions should consider the principle of interoperability when
planning the design of the distributed ledger. Institutions should establish processes to prune the distributed ledger and remove records older than a given number of years or enable processes that allow Data stored on the ledger to be forgotten (i.e. destruction of keys use to encrypt Data).

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only) Business continuity
7.34 Institutions should ensure appropriate business continuity planning with respect to DLT, as it
covers the potential loss of Data and processing capability due to loss of servers or connectivity, and risks such as cyber-crime.
7.35 Institutions should plan, establish and periodically test arrangements to maintain the continuity
of the service/process performed by the DLT Application in the event of an incident that affects the availability of the Application
7.36 Institutions should ensure that their business continuity plan encompasses all the complex
technical areas of DLT, from key storage and key regeneration in the event of catastrophic Data loss to creating new keys when a cyber-crime incident compromises Data security.
7.37 Institutions should consider DLT specific scenarios such as network malfunction or compromise
of Data integrity in their business continuity and disaster recovery plans.
7.38 The Institution’s business continuity plan team should include specialists in DLT and should
monitor cryptographic advances and vulnerabilities such that proactive responses can be developed to avoid system outages.
7.39 In solutions involving public key infrastructure, Institutions should ensure that the business
continuity plan covers the technical integrity of the key generation mechanisms (certificate authorities, hardware security modules etc.), the business processes involved in the secure transportation of the private keys and the authorisation layer around these mechanisms. Customer protection
7.40 Institutions should disclose the governing rules to Customers participating in the distributed
ledger.
7.41 Institutions should disclose any Fees associated with the on-going operation and management of
the distributed ledger including providing notice that Fees may be charged by Third Parties, where applicable.
7.42 Institutions should advise Customers on what they can do to protect their keys from misuse. In
particular, Institutions should inform Customers of the consequences of sharing their private keys and other security information.

Guidelines for Financial Institutions adopting Enabling Technologies (Draft for Discussion Purposes Only)
Section 8: Interpretation
Any clarification or interpretation of the provisions of the Guidelines may be sought from the respective Supervisory Authority:
 FinTech Office of the Central Bank of the UAE (fintechoffice@cbuae.gov.ae);  FinTech Team of the Securities and Commodities Authority (fintech@sca.gov.ae);  Innovation Team at the Dubai Financial Services Authority (innovation@dfsa.ae); and  Financial Technology and Innovation Unit of the Financial Services Regulatory Authority (fintech@adgm.com).