Artificial intelligence and machine learning in

asset management

Barbara Novick
Daniel Mayston Sherry Marcus Rachel Barry
Vice Chairman
Head of Market Co-Head of AI Labs Global Public Policy
Structure and Group
Electronic Trading,
EMEA

Gassia Fox
Bradley Betts Stefano Pasquali Kyle Eisenmann
COO, Global Public
Systematic Active Head of Liquidity BlackRock
Policy Group and
Equity Research, Financial Investment
BlackRock Investment
Modelling Group Stewardship
Stewardship

Background supporting multiple aspects of the investment process.

Technology has become ubiquitous. In 2014, we published Consistent with our view that technology in general can
a ViewPoint titled The Role of Technology within Asset improve the quality and analysis of data for decision-
Management, which documented how asset managers making and drive risk mitigation, we embrace technological
utilize technology in trading, risk management, operations advances, including AI and ML, that can help improve
and client services. As technology continues to evolve and outcomes for our clients.
computing power increases, new use cases are being
identified and new applications are being developed. This These technological tools are part of a larger ecosystem in
applies broadly across sectors, including asset which people make decisions using the information
management. Artificial Intelligence (AI) and Machine generated by computers in various aspects of asset
Learning (ML) are the latest buzzwords in technology management, where a myriad of regulations already apply.
attracting attention. AI and ML reflect the natural evolution These regulations apply regardless of whether a process is
of technology as increased computing power enables performed manually or automated. Specifically, most
computers to sort through large data sets and crunch regulatory regimes across the globe have standards of
numbers to identify patterns and outliers. conduct for trading practices, safety and soundness rules
governing electronic trading, information security
As we laid out in 2014, technology underpins many regulations, disclosure requirements, regulatory reporting,
functions in asset management and has for decades. and regulation regarding the provision of advice.
Virtually all asset managers utilize technology, either
developing their own tools or outsourcing specific In this ViewPoint, we explore the uses of AI and ML in asset
functions to a third party provider. Simply processing large management. While these terms are used frequently, we
quantities of data from portfolio managers, exchanges, find that there are many different understandings of AI and
custodians, rating agencies, and pricing services requires ML; therefore, we begin by defining some of the key terms.
some level of automation to ensure efficiency and Using this foundation, we discuss use cases of AI and ML in
accuracy. the asset management industry, including some specific
use cases at BlackRock. We suggest best practices for
Today, AI and ML are being employed to improve the consideration by asset managers and regulators to factor
customer experience, increase the efficiency and accuracy into their operations and supervision of AI and ML.
of operational workflows, and enhance performance by

The opinions expressed are as of October 2019 and may change as subsequent conditions vary.

blackrock.com/publicpolicy
October 2019 | Public Policy | ViewPoint
GPPGH1019U-988377-1/11
Summary: Best practices and key recommendations
Best Practices: Use of AI/ML in user experiences and interfaces
• The provision of investment advice is heavily regulated. Any new tools or digital advisors are subject to the same
framework of regulation and supervision as traditional advisors, though the applicability and emphasis may vary.

• For AI and ML technologies that have access to client information and sensitive data, it is critical to ensure
robust cybersecurity defenses, including data encryption, cybersecurity insurance, and business continuity
management plans that include incident management frameworks.

Best Practices: Use of AI/ML for operational efficiency

• Investment, operations, and risk professionals should be closely involved in the creation and ongoing oversight
of any model or system that leverages AI or ML, ensuring transparency into the underlying processes used by
the technology.

• Asset managers rely on vast quantities of data, including from external data vendors. Thus, data quality and robust
production monitoring should be of the utmost importance to reduce errors and mitigate operational risks.

• When asset managers choose to buy rather than build AI and ML services and capabilities, clarity on the respective
responsibilities of the third party provider and the asset management firm using the service or tool is essential.
Asset managers should conduct appropriate due diligence on the service providers including ensuring they
have robust testing of the applications, business continuity management, technology disaster recovery
planning, and cybersecurity.

Best Practices: Use of AI/ML in the investment process

• In the design of any model intended to augment human functions, it is critical that the appropriate investment and
risk professionals be closely involved in the creation and ongoing oversight of the technology.

• All data inputs should be robustly tested to ensure models are performing analysis on accurate data sets, and
periodic review procedures should be in place to ensure that no investment process is out-of-date.

• Portfolio managers and risk managers should be able to interpret both the inputs and outputs of the model to
review any investment decision and adjust in new market environments. The more complex the ML technique used
by the model, the higher the risk of obscuring the interpretability of results.

• When used for trading, the AI and ML processes should have appropriate pre-trade controls, development and
release management processes, and real-time ability to monitor and shut off a system. More broadly, such
robust controls should apply to any automated trading process, beyond AI and ML use cases, as electronic
trading should be subject to prudent governance.

Regulatory approaches to the use of AI/ML

• Regulators should balance overseeing the development of new technologies with supporting innovations that
may be beneficial for investors.

• Before pursuing new regulations, we recommend that policy makers consider the applicability of existing regulation
to the uses of AI and ML technologies in asset management and provide additional guidance where appropriate.

• As the applications of new technologies evolve, policy makers should think about how regulation should
similarly evolve and consider providing education and clarification on how existing regulations apply to the use
of new technologies. Many applications of AI and ML are for research purposes only and are not tied to
production processes. Production impact of specific use cases should be considered to determine the appropriate
level of risk and oversight.

• Regulatory sandbox programs can allow for testing of new AI and ML innovations in a controlled environment, and
we encourage regulators to engage with the industry to develop best practices and encourage ongoing innovations.

• Given the global nature of many AI and ML innovations and the financial system as a whole, we encourage
regulators to work together to facilitate globally consistent regimes to ensure that these technologies can
function across borders.

GPPGH1019U-988377-2/11
The proliferation of data tasks ranging from widget assembly on a conveyer belt, to
more complex processes and decision making such as self-
While AI has been around since the 1950s, certain trends
driving cars. Both widget assembly and self-driving cars
have propelled it in the last five to ten years – the growth
follow a methodology that is common across all AI:
of computing processing power, storage, the cloud, and
machines process inputs which subsequently pass through
the proliferation of data. To frame the scale of how much
functions to reach a computer-generated decision as an
data is now available to investors, consider a specific data
output. These functions can be logical (rules -based),
set which is a key component of the modern investment
mathematical, or a combination of both. Consider the
process: detailed financial information about public
example of a “smart home” system that regulates the
companies. The Securities Act of 1933 and Securities
temperature of a room. A user can manually set the
Exchange Act of 1934 require all publicly traded
logical parameters such as the desired temperature and
companies in the U.S. to report universal and verifiable
the time of day to run. What makes the system “smart” is
financial information, including quarterly to annual reports,
that it can be pre-programmed with capabilities that allow
8-K filings, proxy statements, ownership filings and many
the system to change its output, such as automatically
other forms. For the Russell 3000 index, which is
adjusting the temperature of the room according to the
comprised of approximately 3,000 of the largest U.S.
outside temperature, based on a user’s past inputs. It
companies by market capitalization, quarterly and annual
does this without explicit instructions from the user. The
reports alone represent roughly 12,000 documents in a
data inputs in this example are specific and simple. As
given fiscal year. Add to this the availability of transcripts
data sets have increased in size and complexity,
from quarterly earnings calls and investor day presentations,
mathematicians and computer scientists have developed
and there is a trove of data about individual companies that
new techniques to allow for systems to understand more
can be aggregated to identify trends at the sector level.
complex inputs and generate more sophisticated outputs.
The availability of information, such as a company’s These more advanced techniques and models used to
financials and a growing universe of less traditional data analyze varying data sets are known today as ML, which
sets, combined with advances in modern computing, paved we examine in more detail below.
the way for what we refer to as “big data” and new
ML is a specific data science approach to AI. ML programs
technology tools to assess this data.
learn to perform tasks by finding patterns in large data sets
and making inferences instead of following explicit task-
Defining artificial intelligence specific instructions that have already been programmed.
and machine learning Consider recent innovations with virtual assistants such as
Apple’s Siri or Amazon’s Alexa that are already ubiquitous in
The terms “AI” and “ML” are often used interchangeably. millions of households. Based on the sounds a device
While these terms are intertwined, AI is the broader receives (the input), an AI system using ML data science
umbrella term and ML is a subset of AI that reflects the techniques attempts to determine what words have been
evolution of AI.
spoken based on their similarity to a previously captured
AI is the use of machines to replicate human intelligence. audio library. The AI system “learns” as inputs and
This can be thought of on a spectrum ranging from “weak” outcomes, successful or not successful, are recorded and
or “narrow” AI to “strong” AI – with the goal of strong AI stored – large volumes of data allow the underlying models
being replication of intelligence and reasoning. We view AI to recalibrate and better identify patterns. For example, if a
as being in a separate category and distinct from user dictates “Search for spores near me”, Siri will initially
mechanical automation, which is a machine following a set recognize the word “spores”, before quickly modifying the
of pre-defined instructions to accomplish a simple and query to “Search for stores near me”. Apple’s vast
repetitive task. database of other user’s queries has “taught” Siri that it is
far more probable that the user is looking for a storefront
Spectrum of AI rather than a collection of fungi. Siri uses statistical
methods to determine and search a user’s speech input.
Instead of a programmer explicitly telling Siri to make this
Mechanical Wea correction, Siri used vast amounts of data collected to
Strong AI
automation k/ make the rules by itself. It is important to note that data
At present, even the most advanced AI is considered “weak” and models alone are not what allows AI programs like
by the computer science and academic community. Siri to be successful – an entire system comprised of
However, weak AI is still quite powerful; it is used to perform computing architecture, large data storage, mathematical
solutions, and high- performance visualization tools must
all work in tandem to realize the benefits of AI.

GPPGH1019U-988377-3/11
However, many AI/ML applications do not implement an
action and make decisions autonomously like Siri. Rather, User experience and interfaces
the underlying data science models drive a research
Historically, only the wealthy had the access and means
process that enable users to make a more informed
to invest. In recent decades, technology has been the
decision by identifying relationships and patterns in the
driving force in democratizing access to wealth
data that were previously undiscoverable by human efforts
management. For example, an individual investor would
alone. This is akin to GPS navigation applications, such
historically have to contact a stockbroker to facilitate
as Google Maps or Waze, that consume large volumes of
stock transactions, often for high fees. They would need
data such as street mappings, traffic areas, accident reports,
to consult with a tax specialist or accountant to consider
etc. and make recommendations of how to get to the
tax implications and determine how to maximize the value
desired destination. The user makes the end decision of
of these investments.
which path to follow or can choose to deviate from a
recommended path as they see fit based on prior Tools using AI and ML are now available for investors to
experience. The majority of use cases in asset better gain access to the financial markets and receive
management, which we explore throughout the rest of this digital advice, including planning for retirement. Based on
paper, fall within research- oriented applications of AI/ML characteristics such as a user’s age, income, risk tolerance,
rather than production- oriented applications. and desired income in retirement, model-based digital
tools can help investors select the appropriate asset
AI and ML in asset allocation mix to meet their goals. Digital advisors can
incorporate these new tools and provide a way for
management individuals to get personalized investment advice at a lower
Technology has been integrated into asset management for cost than traditional advisory models. Digital advisors may
decades with a variety of uses. Given the evolution of also offer tax-loss harvesting, portfolio allocation
technology broadly in society, it is not surprising that there rebalancing, and digital documentation delivery – features
continue to be new technologies and use cases in asset that previously would have been out of reach without
management. The types of technologies being utilized technological advances like AI and ML.
today – including AI and ML – build on the existing Digital advice can be used to supplement traditional
systems and technology infrastructure. advisors, and most digital advisors offer multiple ways to
Broadly speaking, we break asset management technology engage with a human professional. While some digital
into three main categories: advice firms offer a greater degree of human supervision of
client services and trading systems than others, digital
• User experience and interfaces advisors have a fundamental obligation to oversee their
• Operational efficiency systems and mitigate risks associated with digital
processes. Rather than replacing human advisors, digital
• Investment processes advisors can help them automate processes and more
In each of these areas, technology helps improve efficiency, effectively provide advice at scale. As we outlined in our
manage risk, and enhance decision-making. Importantly, ViewPoint Digital Investment Advice: Robo Advisors Come
all of these technologies involve people and subject of Age, digital advisors are subject to the same framework
matter experts who provide oversight and consider the of regulation and supervision as traditional advisors,
outputs of technologies for more informed decision- though the applicability and emphasis may vary.
making. In the following sections, we take a closer look into
each category, considering AI and ML techniques that can Use at BlackRock: User interfaces
be applied in asset management broadly and how they BlackRock utilizes user interfaces to interact with different
are being deployed specifically within BlackRock. audiences, ranging from large institutions to wealth
managers. These capabilities are supported by over 2,500
experts dedicated to client service, data analytics &
The term asset managers refers to a broad group of production, and technology development. These platforms
firms and investment strategies. For purposes of this are built to service the highly custom financial needs of
paper, we are focused on traditional asset managers advisors at scale to seek to create the best financial
which we differentiate from nontraditional asset outcomes for clients. Currently, we are exploring the use of
managers (including hedge funds, private equity, and AI/ML in our tools with the aim of enhancing our ability to
other alternative investment strategies). In our connect with clients and tailor solutions to help them
experience, traditional asset managers employ AI and achieve their financial goals.
ML to varying degrees. This paper focuses on common
uses; this is not meant to be an exhaustive list.

GPPGH1019U-988377-4/11
Examples of BlackRock tools to improve the user experience and interfaces

Solution Overview

Aladdin Wealth provides an institutional-quality solution to wealth

management firms and banks, includingenterprise-wide business and risk
oversight and portfolio analysis and construction capabilities

FutureAdvisor digitizes the wealth management process for financial

institutions and their advisors to onboardand serve new client segments in a
scalable way

iRetirehelps financial professionals build plans thatare focusedonretirement

saving, spendingandincome quickly andefficiently usinga repeatable process
– so they can focus onclientserviceand assetacquisition

Advisor Center provides practice management capabilities, like scenario

analysis and tax insights, powered by Aladdin, on the web

References to BlackRock’s user interfaces are intended to illustrate how asset managers are able to use technology in various ways and are not intended to promote
any specific BlackRock asset management service or technology solution.

vast quantities of data on financial instruments that asset

managers rely on. Data quality is of the utmost importance
Best Practices: Use of AI/ML in as making fewer errors reduces operational risk and
user experiences and interfaces protects the end clients. Data may be missing, stale, or may
• The provision of investment advice is heavily contain other errors. To help alleviate this, machine
regulated. Any new tools or digital advisors learning techniques can be used to identify and flag
are subject to the same framework of outliers based on statistical assessments. For example, a
regulation and supervision as traditional model might be able to take known inputs such as the
advisors, though the applicability and average price of a stock to determine if the latest price
emphasis may vary. received from a vendor is erroneous.

• For AI and ML technologies that have access to Asset managers have always used technology to drive
client information and sensitive data, it is critical efficiencies and create better outcomes for investors, and
to ensure robust cybersecurity defenses, newer technologies like AI and ML expand an asset
including data encryption, cybersecurity manager’s toolkit to do so even more effectively.
insurance, and business continuity management
plans that include incident management Use at BlackRock: Operational efficiency
Data scientists at BlackRock leverage AI and ML to scale
Operational efficiency data tasks for operational processes. Use cases include
surveillance, data cleansing, and support functions. For
Asset managers have a fiduciary duty to servicing clients
example, we create and review over 1 million daily risk and
at a lower cost. Technology has been a driving force to
exposure reports on portfolios through our system called
help streamline and manage the processing of internal
Aladdin. These reports are generated from vast data sets.
data and external data feeds, and the post-trade
Because most data errors are systematic in nature (e.g.,
operational processes from confirmation, settlement of
trades to reconciliation of positions, cash balances, and data is missing, out of date, or incorrect), data quality
net asset values (NAVs). control (QC) and cleansing makes the process ideal for our
data teams to use AI/ML models.
A key application of AI/ML to operational functions is the
quality checking, monitoring, and exception handling of
the

GPPGH1019U-988377-5/11
BlackRock’s daily production process for report creation

1MM+ daily risk and exposure reports for

Aladdin engine
Data sources

• Smart beta portfolios that can use various lenses to

decide how to weight allocations within an index to
Best Practices: Use of AI/ML for favor certain characteristics, such as sustainable
operational efficiency dividends or low volatility.
• Investment, operations, and risk professionals • Index investing and ETFs which facilitate efficient and
should be closely involved in the creation and accurate replication of an index and manage the
ongoing oversight of any model or system that mechanical trading processes that deliver index
leverages AI or ML, ensuring transparency into replication. Index funds and ETFs have been made
the underlying processes used by the possible by innovations in product structure
technology. supported by technology.
• Asset managers rely on vast quantities of data,
including from external data vendors. Thus, data Research and alpha generation:
quality and robust production monitoring should seeking to generate superior returns
be of the utmost importance to reduce errors
and mitigate operational risks.
for clients
To generate higher returns for actively managed portfolios,
• When asset managers choose to buy rather
asset managers study and interpret the available data on
than build AI and ML services and capabilities,
the assets in which they invest on behalf of clients. In the
clarity on the respective responsibilities of the
case of equities, this process might begin with basic
third party provider and the asset management
financial information provided by company filings.
firm using the service or tool is essential. Asset
However, there is now a broad range of other information
managers should conduct appropriate due
that can signal a company’s future performance. Asset
diligence on the service providers including
managers have developed AI and ML tools to compile,
ensuring they have robust testing of the
cleanse, and analyze the universe of data available,
applications, business continuity management,
including analyst reports, macroeconomic data (e.g., GDP
technology disaster recovery planning, and
growth, unemployment) as well as newer “alternative” data

Investment process sources. Examples of alternative data include GPS and

satellite imagery to see where consumers are going,
Technology is used to facilitate multiple aspects of the internet searches and tweets to see what people are
investment process including the data and research researching and talking about, and employee satisfaction
processes that drive the creation of alpha signals and data, all of which can be accessed online today. These data
models, pre-trade analysis, and understanding investment points can help portfolio managers better assess individual
risks in a given portfolio. companies and sectoral trends. As the volume of real-time
Some examples of how technology is used by different data available increases beyond the capacity of individuals
investment strategies include: to analyze and understand it, the ability to compile, cleanse,
and evaluate that data is increasingly important. AI and ML
• Mathematical models that can analyze datasets (both enable asset managers to find patterns in this data at scale,
limited and large, traditional and “alternative”) to potentially identifying signals for generating returns for
identify patterns and insights as inputs to the clients.
investment decision-making process.

GPPGH1019U-988377-6/11
 The review process involves human oversight at every step,
including peer reviews, research board approvals,
Best Practices: Use of AI/ML production coding by an independent technology team and
in research and alpha final sign off by a researcher. Once it is applied to a
generation portfolio, the investment model is monitored by both the
portfolio manager and an independent risk management
• In the design of any model intended to
team to ensure it is behaving as expected. Portfolio
augment human functions, it is critical that the
managers have ultimate visibility and control of their
appropriate investment, trading and risk
models, and they are able to perform “off-model
professionals be closely involved in the
overrides” if a particular investment does not make sense
creation and ongoing oversight of the
in the current market context.
technology.

• All data inputs should be robustly tested to Trading processes: minimizing

ensure models are performing analysis on
transaction costs and market impact
accurate data sets, and periodic review
While analyzing data to generate investment ideas is a core
procedures should be in place to ensure that
part of the investment process, equally crucial is the
no investment process is out- of-date.
implementation of the idea through efficient trade
• Portfolio managers and risk managers should be execution. Markets have experienced material
able to interpret both the inputs and outputs of transformation over the past few decades due to advances
the model to review any investment decision and in technology and the adoption of new regulation. These
adjust in new market environments. The more drivers have increased the speed and automation of
complex the ML technique used by the model, markets and fostered the emergence of new venues and
Use at BlackRock: Systematic Active Equity market participants. As traditional over the counter
execution has largely given way to electronic trading across
BlackRock’s Systematic Active Equity team leverages a
primary asset classes, this has increased the breadth of
rigorous research process that combines human insights
digital touchpoints in the investment lifecycle, generating a
with data, technology and mathematics to seek better
vast dataset for trading teams to analyze in order to
returns for our clients. Each of the inputs and outputs of
minimize costs. Consider the example of a basic trade
an investment model undergo a rigorous testing and
order, illustrated on the next page.
approval process before the model is released into
production.

Example of how we analyze large data sets to identify signals

Using text analysis techniques to anticipate future changes to company earnings guidance

Analyze Measure

Use technology
Transform
to analyze over
unstructured text
5,000 earnings
into proprietary
call transcripts
measures of
every quarter
trending analyst
and more than
sentiment
6,000 broker
reports every
day
Traditional approach:
Positive
Individual reports read by hand – or
Negative await
analyst revision to occur

GPPGH1019U-988377-7/11
Basic order and trade execution
Brokers
Broker A
$3/share

Marketplace
“Buy 100 PortfolioRaises Trading desk Receives & routesBroker
order B
$4/share Facilitate
shares of managerorder Buyers &
transaction
Company sellers
Broker C
$5/share

If a portfolio manager decides to buy 100 shares of

trading and its integration into trading decisions to improve
Company XYZ and the trade is approved, it is up to the
execution for clients (which we discuss in our ViewPoint
trading function to accomplish three goals:
titled U.S. Equity Market Structure: An Investor
1. Buy the shares at the lowest price possible; Perspective).

2. Pay as little in facilitation fees as possible to minimize

the cost of the transaction;

3. Minimize the market impact of the transaction. Best

Use Practices:
at BlackRock: Userisk
Liquidity of AI/ML in
management
AI and ML models can assist human traders in selecting trade execution
BlackRock has a financial modeling group that oversees
the routing destination or execution methodology which developing investment risk models at the security and
• In the design of any model intended to augment
portfolio level. Last year, we developed a liquidity risk model
optimizes for the three objectives above. For example, a human functions, it is critical that the appropriate
that leverages ML techniques to estimate average daily
trading model can analyze broker-dealer inventory, investment, trading and risk professionals be
volumes within the transaction cost framework. Another
pricing data, and historical transactions to make closely involved in the creation and ongoing
liquidity model still in a research phase can utilize ML to
recommendations on which counterparties or venues oversight of the technology.
estimate redemption risk in our mutual funds.
would result in better trading performance. These
• When used for trading, the AI and ML processes
recommendations can then be incorporated using a Our financial modelers often choose to leverage more
should have appropriate pre-trade controls,
programmed instruction set (typically not using AI and ML simplistic models in favor of more sophisticated ones that
building and release management processes,
technology) within a tool called a “smart order router,” have higher interpretability – the ability to attribute the
and real-time ability to monitor and shut off a
which can display suggestions to traders or send orders output of the model to the inputs. This gives us full control
system. As the primary usage of AI and ML in
out electronically. and understanding of how the model made forecasts. We
trading processes
believe that more complex ML techniques that are less
Other parts of the trading process where AI/ML has been to conduct research, these controls
interpretable should only be used after more interpretable
techniques may be applicable include external execution should apply to automation of trading more
models have been explored and the performance increase
tools like broker-dealer execution algorithms. It is broadly beyond AI and ML.
has business justifications.
important to delineate the relevant responsibilities across
the trading lifecycle, particularly as it relates to these third
party processes where the asset manager does not
develop the automated execution algorithms. In this
case, the asset manager is submitting instructions to an
external broker- dealer execution algorithm, which then
facilitates the transaction for a fee. While it is important
for the asset manager to understand the execution
methodology of the third party, the actual implementation
is managed externally.

Use at BlackRock: Trade execution

Our data scientists leverage AI and ML to analyze
BlackRock’s trading data captured electronically to help
identify patterns in transaction costs. Based on these
patterns, our traders can make more informed decisions
on how to route trades for execution to reduce latency,
cost, and market impact. BlackRock has a dedicated
electronic trading and market structure team responsible
for designing and overseeing the shift towards electronic

GPPGH1019U-988377-8/11
Risk management and • Technology risk: RQA’s Technology Risk team works
with all technology functions to help ensure technology
oversight at BlackRock controls are in place to protect clients and the firm, for
BlackRock was founded on the principle that risk example by monitoring system resilience and
management is a key tenet of asset management. We take performance, or assessing software development and
a multi-faceted approach to risk management embedding testing practices - aligning practices with industry and
checks and balances and instilling a culture of risk regulatory standards.
management across portfolio management, trading and
• Third party risk: RQA’s Third Party risk team helps
operations functions. As a first line of defense, the firm
to identify and manage any risks that may arise from
expects individual portfolio management, research and
third party vendors by conducting due diligence on
trading teams to take primary responsibility for managing
potential new vendors, validating the appropriateness
the investment risks including liquidity risk management
of vendor controls and identifying potential risks prior
associated with the portfolios that they manage, and to
to on- boarding and throughout the life of significant
ensure that they are following key controls, fund
relationships. This includes any potential risks arising
mandates, and regulations.
from the use, by third parties, of AI and ML
The second line of defense is our dedicated risk applications in their provision of services to
management group, called Risk and Quantitative Analysis BlackRock.
(RQA), which monitors the risk profiles of portfolios
managed on behalf of clients and regularly engages with • Model risk: RQA’s model risk team has developed
the portfolio management teams to ensure risk positioning standards and policies for the design, implementation
is deliberate, diversified, and scaled. BlackRock’s RQA and use of quantitative models for investment, risk
function is independent from the investment teams and management, and valuation purposes. These include
monitors various aspects of the investment management requirements for documentation and testing,
process, including investment risks, operational risks, independent model validation, release controls,
technology risks, third party risk and model risks. This performance monitoring and governance. Such
helps to ensure that each of these risks is clearly requirements are readily applicable to, and can
identified, assessed, managed and monitored. These risk mitigate risk in, situations where models incorporate
programs described below are inclusive of AI and ML AI and ML.
applications and provide additional oversight as a The third line of defense is our internal audit function that
supplement to the risk management undertaken within independently validates investment businesses’ adherence
each applicable business function: to key controls and policies and provides independent
• Operational risk: RQA’s Operational Risk team works substantiation of control issues either self-identified by
with all operating and investment teams to proactively investment teams or flagged by independent risk
identify risk and assess the adequacy of controls. management. Internal audit independently assesses and
While more automated processes can reduce risk, validates the scope and efficacy of RQA’s risk management
consideration is given to the design and performance activities described above.
of such processes as part of operational risk Finally, the Enterprise Risk Management Committee
assessment. The team is also engaged whenever (ERMC), which includes leaders across the firm’s risk
operating events or errors occur in any investment or management and technology functions as well as members
operational process, to ensure that root causes are of the Global Executive Committee, plays a key oversight
quickly and clearly identified and controls improved. role. The ERMC looks to identify the material risks of the
firm, understand and define the firm’s risk tolerance, assure
each identified risk has an owner, review risk mitigation
Enterprise risk oversight at BlackRock activities, and communicate individual, aggregated, and
emerging risks to BlackRock’s Board of Directors.

Enterprise Risk Management Committee

Operational Risk Technology Risk Third Party Risk Model Risk

Management Management Management Management

GPPGH1019U-988377-9/11
Regulatory approaches to the innovations and the financial system as a whole, we
encourage regulators to work together to facilitate globally
use of AI/ML consistent regimes to ensure that these technologies can
Regulators across the globe are closely following the function across borders.
development and uses of technology in asset management
while also considering how technology, including AI and Conclusion
ML, can improve regulatory functions. Before pursuing new
Technology will continue to play an integral role across
regulations, we recommend that policy makers consider the
various asset management functions as it has for decades.
applicability of existing regulation to new technologies and
As new tools are developed, computing power becomes
provide additional guidance where appropriate. The use and
more affordable, and the availability of data continues to
development of AI/ML is not stagnant, and policymakers
increase, additional use cases for AI and ML in asset
should think about how regulation should similarly evolve.
management will emerge. We are already seeing tools that
We recommend policy makers consider providing education
can help mitigate risk, reduce costs, generate better returns,
and clarification on how existing regulations apply to use
and deliver products and services more efficiently for
cases of these technologies.
clients.
Many applications of AI and ML are for research purposes
Appropriate controls and direct oversight by human
only and not tied to production processes. Production
professionals are critical to ensure the success of AI and ML
impact of specific use cases should be considered to
applications. Some firms, including BlackRock, have
determine the appropriate level of risk and oversight. The
implemented a governance structure that provides
use of regulatory sandbox programs can allow for testing of
additional oversight. Regardless of the approach a firm
new AI and ML innovations in a controlled environment, and
deploys, human judgement, review, and robust testing are
we encourage regulators to engage with the private sector
key. We encourage regulators and the industry to work
to develop best practices and encourage ongoing
together as these technologies evolve to provide better
innovations. Given the global nature of many AI and ML
outcomes for investors.

Related Content
For access to our full collection of public policy commentaries, including the ViewPoint series and
comment letters to regulators, please visit www.blackrock.com/corporate/en-us/insights/public-policy.
ViewPoint – The Role of Technology Within Asset Management
ViewPoint – The Role of Third Party Vendors in Asset Management
ViewPoint – Digital Investment Advice: Robo Advisors Come of Age
ViewPoint – Mark-to-market structure: An end-investor perspective on the evolution of developed equity markets
Remarks by Barbara Novick – The Future for Financial Services

10

GPPGH1019U-988377-10/
This publication represents the regulatory and public policy views of BlackRock. The opinions expressed herein are as of October 2019 and are subject to change at any time due to
changes in the market, the economic or regulatory environment or for other reasons. The information herein should not be cons trued as sales material, research or relied upon in making
investment decisions with respect to a specific company or security. Any reference to a specific company or security is for i llustrative purposes and does not constitute a
recommendation to buy, sell, hold or directly invest in the company or its securities, or an offer or invitation to anyoneto invest in any fun ds, BlackRock or otherwise, in any jurisdiction.
There is no guarantee that any forecasts madewill come to pass. Reliance upon information in this material is at the sole discretion of the reader.

In the U.S., this material is available for public distribution. In the UK, issued by BlackRock Investment Management (UK) Limited ( authorised and regulated by the Financial Conduct
Authority). Registered office: 12 Throgmorton Avenue, London, EC2N 2DL. Registered in England No. 2020394. Tel: 020 7743 3000 . For your protection, telephone calls are usually
recorded. BlackRock is a trading name of BlackRock Investment Management (UK) Limited. This material is for distribution to Professional Clients (as defined by the FCA Rules) and
Qualified Investors and should not be relied upon by any other persons. In the EEA, issued by BlackRock (Netherlands) BV: Amstelplein 1, 1096 HA, Amsterdam, Tel: 020 – 549 5200,
Trade Register No. 17068311. BlackRock is a trading name of BlackRock (Netherlands) BV. For qualified investors in Switzerland, this material shall be exclusively made available to,
and directed at, qualified investors as defined in the Swiss Collective Investment Schemes Act of 23 June 2006, as amended. In Australia, issued by BlackRock Investment
Management (Australia) Limited ABN 13 006 165 975, AFSL 230 523 (BIMAL). This material is not a securities recommendation or an offer or solicitation with respect to the purchase
or sale of any securities in any jurisdiction. The material provides general information only and does not take into account your individual objectives, financial situation, needs or
circumstances. Before making any investment decision, you should therefore assess whether the material is appropriate for you and obtain financial advice tailored to you having regard
to your individual objectives, financial situation, needs and circumstances. BIMAL, its officers, employees and agents believe that the informat ion in this material and the sources on
which it is based (which may be sourced from third parties) are correct as at thedate of publication. While every care has been taken in the preparat ion of this material, no warranty of
accuracy or reliability is given and no responsibility for the information is accepted by BIMAL, its officers, employees or agents. Any investment is subject to investment risk, including
delays on the payment of withdrawal proceeds and the loss of income or the principal invested. While any forecasts, estimates and opinions in this mat erial are made on a reasonable
basis, actual future results and operations may differ materially from the forecasts, estimates and opinions set out in this material. No guarantee as to the repayment of capital or the
performance of any product or rate of return referred to in this material is made by BIMAL or any entity in the BlackRock group of companies. In Singapore, this is issued by BlackRock
(Singapore) Limited (Co. registration no. 200010143N). In Hong Kong, this material is issued by BlackRock Asset Management North Asia Limited and has not been reviewed by the
Securities and Fu tures Commission of Hong Kong. In South Korea, this material is for distribution to the Qualified Professional Investors (as defined in the Financial Investment
Services and Capital Market Act and its sub-regulations). In Taiwan, independently operated by BlackRock Investment Management (Taiwan) Limited. Address: 28F., No. 100, Songren
Rd., Xinyi Dist., Taipei City 110, Taiwan. Tel: (02)23261600. In Japan, this is issued by BlackRock Japan. Co., Ltd. (Financial Instruments Business Operator: The Kanto Regional
Financial Bureau.License No375, Association Memberships: Japan Investment Advisers Association, the Investment Trusts Association, Japan, Japan Securities Dealers Association,
Type II Fina ncial Instruments Firms Association.) For Professional Investors only (Professional Investor is defined in Financial Instruments and Exchange Act). In China, this material
may not be distributed to individuals resident in the People's Republic of China ("PRC", for such purposes, excluding Hong Kong, Macau and Taiwan) or entities registered in the PRC
unless such parties havereceived all the required PRC government approvals to participate in any investment or receive any investment advisory or investment management services. For
Other APAC Countries, this material is issued for Institutional Investors only (or professional/sophisticated /qualified investors, as such term may apply in local jurisdictions) and does
not constitute inves tment advice or an offer or solicitation to purchase or sell in any securities, BlackRock funds or any investment strategy nor shall any securities be offered or sold to
any person in any jurisdiction in which an offer, solicitation, purchase or sale would be unlawful under the securities laws of such jurisdiction. In Latin America, for institutional investors
and financial intermediaries only (not for public distribution). This material is for educational purposes only and does not constitute investment advice or an offer or solicitation to sell or a
solicitation of an offer to buy any shares of any fund or security and it is your responsibility to inform yourself of, and to observe, all applicable laws and regulations of your relevant
jurisdiction. If any funds are mentioned or inferred in this material, such funds may not be registered with thesecurities regulators in any Latin American country and thus, may not
bepublicly offered in any s uch countries. Thesecurities regulators of any country within Latin America have not confirmed the accuracy of any information contained herein. Investing
involves risk, including possibl e loss of principal. The contents of this material are strictly confidential and must not be passed to any third party. In Mexico if any funds, securities or
investment strategies are mentioned or inferred in this material, such funds, securities or strategies have not been registered with the National Banking and Securities Commission (
Comisión Nacional Bancaria y de Valores, the “CNBV”) and thus, may not bepublicly offered in Mexico. The CNBV has not confirmed the accuracy of any information contained herein.
The provision of investment management and investment advisory services is a regulated activity in Mexico, subject to strict rules, and performed under the supervision of the CNBV.
BlackRock Mexico, S.A. de C.V., Asesor en Inversiones Independiente(“BLKMX”) is a Mexican subsidiary of BlackRock, Inc., registered with the CNBV as an independent investment
advisor under registration number 30088-001-(14085)-20/04/17, and as such, authorized to provide Investment Advisory Services. BlackRock México Operadora, S.A. de C.V., Sociedad
Operadora de Fondos de Inversión (“BlackRock MX Operadora” and together with BLKMX, “BlackRock México”) are Mexican subsidiaries of BlackRock, Inc., authorized by the CNBV. For
more information on theinvestment services offered by BlackRock Mexico, please review our Investment Services Guide available in www.BlackRock.com/mx. Reliance upon information
in this material is at your sole discretion. Blac kRock México is not authorized to receive deposits, carry out intermediation activities, or act as a broker dealer, or bank in Mexico.
Further, BlackRock receives revenue in the form of advisory fees for our mutual funds and exchange traded funds and management fees for our collective investment trusts.

©2019 BlackRock, Inc. All Rights Reserved. BLACKROCK is a registered trademark of BlackRock, Inc. All other trademarks are th oseof their respective owners.

GPPGH1019U-988377-11/