没有合适的资源?快使用搜索试试~ 我知道了~
金融中的人工智能、机器学习和偏见:走向负责任的创新.docx
1.该资源内容由用户上传,如若侵权请联系客服进行举报
2.虚拟产品一经售出概不退款(资源遇到问题,请及时私信上传者)
2.虚拟产品一经售出概不退款(资源遇到问题,请及时私信上传者)
版权申诉
0 下载量 45 浏览量
2024-04-23
07:20:23
上传
评论
收藏 125KB DOCX 举报
温馨提示
试读
32页
金融中的人工智能、机器学习和偏见:走向负责任的创新.docx
资源推荐
资源详情
资源评论
Fordham Law Review
Volume 88 Issue 2 Article 5
2019
Artificial Intelligence, Machine Learning, and Bias in Finance:
Toward Responsible Innovation
Kristin Johnson
Tulane
University
Law
School
Frank
Pasquale
University of Maryland Francis King Carey School of Law
Jennifer Chapman
University of Maryland Francis King Carey School of Law
Follow this and additional works at: https://ir.lawnet.fordham.edu/flr
Part of the Banking and Finance Law Commons, Internet Law Commons, Law and Society Commons,
and the Science and Technology Law Commons
Recommended
Citation
Kristin Johnson, Frank Pasquale, and Jennifer Chapman, Artificial Intelligence, Machine Learning, and
Bias in Finance: Toward Responsible Innovation, 88 Fordham L. Rev. 499 (2019).
Available at: https://ir.lawnet.fordham.edu/flr/vol88/iss2/5
This Symposium is brought to you for free and open access by FLASH: The Fordham Law Archive of Scholarship
and History. It has been accepted for inclusion in Fordham Law Review by an authorized editor of FLASH: The
Fordham Law Archive of Scholarship and History. For more information, please contact
tmelnick@law.fordham.edu.
ARTIFICIAL INTELLIGENCE, MACHINE
LEARNING, AND BIAS IN FINANCE: TOWARD
RESPONSIBLE INNOVATION
Kristin Johnson,* Frank Pasquale** & Jennifer Chapman***
INTRODUCTION
Over the last decade, a growing number of digital startups launched bids
to lure business from the financial services industry.
1
Armed with what they
claim are vast quantities of data and sophisticated algorithmic platforms
capable of interpreting the data,
2
these financial technology (“fintech”)
3
* McGlinchey Stafford Professor of Law, Gordon Gamm Faculty Fellow, Tulane University
Law School.
** Piper & Marbury Professor of Law, University of Maryland Francis King Carey School
of Law; Affiliate Fellow, Yale Information Society Project.
*** Ryan H. Easley Research Fellow, University of Maryland Francis King Carey School of
Law. This Essay was prepared for the Symposium entitled Rise of the Machines: Artificial
Intelligence, Robotics, and the Reprogramming of Law, hosted by the Fordham Law Review
and the Neuroscience and Law Center on February 15, 2019, at Fordham University School
of Law. We would like to thank Kathleen Engel for her comments on the draft.
1. Andrew Ross Sorkin, Fintech Firms Are Taking On the Big Banks, but Can They
Win?, N.Y. TIMES:
DEALBOOK (Apr. 6, 2016), https://www.nytimes.com/2016/04/07/
business/dealbook/fintech-firms-are-taking-on-the-big-banks-but-can-they-win.html
[https://perma.cc/8Z6C-DWKQ]; The Fintech Revolution, ECONOMIST (May 9, 2015),
https://www.economist.com/news/leaders/21650546-wave-startups-changing-financefor-
better-fintech-revolution [https://perma.cc/UWF3-Z7PZ].
2. FTC,
BIG
DATA: A
TOOL
FOR
INCLUSION
OR
EXCLUSION?,
at
i
(2016),
https://www.ftc.gov/system/files/documents/reports/big-data-tool-inclusion-or-exclusion-
understanding-issues/160106big-data-rpt.pdf [https://perma.cc/R44Z-P4P5].
3. In previous publications, Frank Pasquale has examined “incrementalist” fintech,
which utilizes technology to provide standard financial services, and “futurist” fintech, in
which the entire financial system is remade due to distributed technologies. See Exploring the
Fintech Landscape: Hearing Before the S. Comm. on Banking, Hous. & Urban Affairs, 115th
Cong. (2017) (statement of Frank Pasquale, Professor of Law, University of Maryland). In
this Essay, we use the term “fintech firms” to refer to nondepository financial services firms
that integrate artificial intelligence technology and predictive analytics into their business
models. We acknowledge that, while there is no universally adopted definition for the term
“fintech,” many use the term as a catchall for a broader group of financial services firms that
integrate a diverse body of technologies and engage in digital transfers, storage, payments
systems, and lending, as well as the origination of virtual currency and robo-advising. See,
e.g., Rory Van Loo, Making Innovation More Competitive: The Case of Fintech, 65 UCLA
L. REV. 232, 238–40 (2018).
499
500
FORDHAM LAW REVIEW
[Vol. 88
firms have revived long-standing debates regarding the architectural design,
4
regulatory framework,
5
and role of the financial services industry.
6
Financial product developers and financial service providers have long
engaged statistical and probability models as well as predictive analytics to
forecast performance.
7
So fintech is not entirely new. However, sometimes
a change in quantity can amount to a change in quality. That may be
happening in fintech now, as the inclusion of increasingly comprehensive
databases, as well as new methods of analysis, means that many fintech firms
deploy extremely complex algorithms (including assemblages of earlier
models) to predict the likelihood of repayment and profitability of
customers.
8
According to some futurists, financial markets’ automation will
substitute increasingly sophisticated, objective, analytical, model-based
assessments of, for example, a borrower’s creditworthiness for direct human
evaluations irrevocably tainted by bias and subject to the cognitive limits of
the human brain.
9
However, even if they do occur, such advances may
violate other legal principles.
10
Consider, for example, the application of learning algorithms in credit
markets. Some fintech firms aim to adapt learning algorithms to consider
nontraditional data in assessing creditworthiness
11
and claim that they will
integrate historically excluded individuals into credit markets and expand
4. See infra Part I.A.
5. See infra Part II.A.
6. See, e.g., MEHRSA BARADARAN, HOW THE OTHER HALF BANKS: EXCLUSION,
EXPLOITATION, AND THE THREAT TO DEMOCRACY (2015) (arguing for a “public option” in
consumer banking); E. GERALD CORRIGAN, FED. RESERVE BANK OF MINNEAPOLIS, ARE BANKS
SPECIAL?: ANNUAL REPORT 1982 (1982), https://www.minneapolisfed.org/publications/
annual-reports/ar/annual-report-1982-complete-text [https://perma.cc/9V29-W9V8] (raising
fundamental questions regarding the role of banks and discussing their prudential regulation).
7. ANTHONY SAUNDERS & MARCIA CORNETT, FINANCIAL INSTITUTIONS MANAGEMENT:
A RISK MANAGEMENT APPROACH 97–103 (9th ed. 2017).
8. See infra Part I. Note that customers who are late with payments may be much more
profitable than a traditionally good credit risk, since they will be paying more in interest and
fees.
9. OECD, FINANCIAL MARKETS, INSURANCE AND PENSIONS: DIGITALISATION AND
FINANCE 10–13 (2018), https://www.oecd.org/finance/private-pensions/Financial-markets-
insurance-pensions-digitalisation-and-finance.pdf [https://perma.cc/GW8N-RPXL].
10. See, e.g., Odia Kagan, Finnish DPA Orders Company to Modify Automated
Creditworthiness Assessment, Improve Disclosures, FOX ROTHSCHILD (Apr. 27, 2019),
https://dataprivacy.foxrothschild.com/2019/04/articles/european-union/finnish-dpa-orders-
company-to-modify-automated-creditworthiness-assessment-improve-disclosures/
[https://perma.cc/3K55-8MXN] (reporting that the Finnish Data Protection Authority ordered
a firm to “provide individuals with information on the logic behind the decision-making
process, its relevance to the credit decision and its consequences for the borrower” pursuant
to the General Data Protection Regulation’s provisions guaranteeing a right to an explanation).
11. Examining the Use of Alternative Data in Underwriting and Credit Scoring to Expand
Access to Credit: Hearing Before the H. Comm. on Fin. Servs., 116th Cong. 4–6_(2019)
(statement of Kristin N. Johnson, McGlinchey Stafford Professor of Law and Associate Dean
of Faculty Research, Tulane University Law School); see also Julapa Jagtiani & Catharine
Lemieux, The Roles of Alternative Data and Machine Learning in Fintech Lending: Evidence
from the Lending Club Consumer Platform (Fed. Reserve Bank of Phila., Working Paper No.
18-15, 2018), https://www.philadelphiafed.org/-/media/research-and-data/publications/
working-papers/2018/wp18-15r.pdf [https://perma.cc/2TT9-U6A3].
2019]
TOWARD RESPONSIBLE INNOVATION
501
access to credit to the thirty-three million unbanked and underbanked
households in the United States,
12
as well as the nearly two billion
individuals and families globally who lack access to financial services
13
—a
group disproportionately composed of women and people of color.
14
How might fintech firms accomplish such a lofty goal? Early fintech firms
promising to better integrate underresourced communities into financial
services markets typically introduced digital money transfer services that
facilitated cash distributions among users (such as PayPal, Apple Pay, or
Venmo)
15
and credit platforms that offered digitally distributed consumer
loans. Money transmission services can provide vital peer-to-peer platforms
for those who lack access to conventional bank branches or personal
checking and savings accounts. Because credit is an indisputably important
resource for low-income families in smoothing consumption
16
and creating
economic stability,
17
evaluating the integration of automated decision-
making algorithms in credit markets raises underexplored normative
concerns including the transparency and accountability obligations of fintech
firms, the social welfare effects of permitting fintech firms to operate in credit
markets, and the necessity of effective state and federal supervision of fintech
firms’ pricing (interest rates), marketing techniques, and structuring of credit
products.
With a few quick taps on a smart phone, consumers can access a growing
universe of apps that offer discounted interest rates on consumer loans to
borrowers with “near prime,” “subprime,” and well-below subprime credit
12. FDIC, 2017 FDIC NATIONAL SURVEY OF UNBANKED AND UNDERBANKED
HOUSEHOLDS 1 (2018), https://www.fdic.gov/householdsurvey/2017/2017report.pdf
[https://perma.cc/F7TP-SMHZ] (indicating that 6.5 percent of U.S. households (or 8.4 million
households) were unbanked in 2017 and 18.7 percent of U.S. households (24.2 million) were
underbanked, meaning that the household had a checking or savings account but also obtained
financial products and services outside of the banking system).
13. ASLI DEMIRGÜÇ-KUNT ET AL., WORLD BANK GRP., THE GLOBAL FINDEX DATABASE
2017: MEASURING FINANCIAL INCLUSION AND THE FINTECH REVOLUTION 4 (2018),
http://documents.worldbank.org/curated/en/332881525873182837/pdf/126033-PUB-
PUBLIC-pubdate-4-19-2018.pdf [https://perma.cc/YHT7-R7P8] (indicating that “about 1.7
billion adults remain unbanked—without an account at a financial institution or through a
mobile money provider”).
14. See generally Louise Seamster, Black Debt, White Debt, CONTEXTS, Winter 2019, at
30.
15. Adam Levitin, Pandora’s Digital Box: The Promise and Perils of Digital Wallets,
166 U. PA. L. REV. 305, 335 (2018).
16. Brian T. Melzer, The Real Costs of Credit Access: Evidence from the Payday Lending
Market, 126 Q.J. ECON. 517, 522 (2011) (indicating that loans give families flexibility “in
managing consumption over time” yet may create “substantial debt service burdens”).
17. See, e.g., Christine L. Dobridge, For Better and for Worse?: Effects of Access to High-
Cost Consumer Credit (Fed. Reserve Bd., Working Paper No. 2016-056, 2016),
https://www.federalreserve.gov/econresdata/feds/2016/files/2016056pap.pdf
[https://perma.cc/7PFP-G5VC] (“[P]ayday credit access improves well-being for households
in distress by helping them smooth consumption. In periods of temporary financial distress—
after extreme weather events like hurricanes and blizzards—I find that payday loan access
mitigates declines in spending on food, mortgage payments, and home repairs. In an average
period, however, I find that access to payday credit reduces well-being.”).
502
FORDHAM LAW REVIEW
[Vol. 88
scores.
18
For proponents, the launch of fintech firms marks a new frontier in
the ever-expanding utopian vision of the “technological sublime” or faith-
like devotion to the potential for technology to transform us into a more
equitable and just society.
19
Consumer advocates are justifiably skeptical. While legally prohibited
today, well-documented discriminatory,
20
exclusionary, and predatory credit
market practices persist.
21
In light of creditors’ history of exploiting
unbanked and underbanked communities, even fintech firms’ plans for
greater inclusion demand careful scrutiny.
Consider the disturbing tales emerging of digital debt platforms peddling
payday loan–style arrangements masked by the opaque and unassailable
shroud of innovation and financial inclusion.
22
Kevin Donovan and Emma
Park share harrowing narratives of aggressive marketing campaigns by text
message that entice borrowers already consumed by “perpetual debt” to
borrow at expensive, ballooning interest rates.
23
Further, in the event that
they fail to repay the loans, some fintech firms harass overextended
borrowers with incessant and embarrassing payment alerts on their mobile
phones.
24
Cash-strapped borrowers who lack the resources to meet their
daily expenses enter a downward spiral of indebtedness. Borrowers on the
18. Sarah McBride, ZestFinance Targets Loans at the Cream of Subprime Borrowers,
REUTERS (July 14, 2015), https://www.reuters.com/article/us-venture-zestfinance-subprime/
zestfinance-targets-loans-at-the-cream-of-subprime-borrowers-idUSKCN0PP0B320150715
[https://perma.cc/74BK-SPG8].
19. See generally VINCENT MOSCO, THE DIGITAL SUBLIME: MYTH, POWER, AND
CYBERSCAPE (2004); DAVID E. NYE, AMERICAN TECHNOLOGICAL SUBLIME (1994).
20. Louise Seamster & Raphaël Charron-Chénier, Predatory Inclusion and Education
Debt: Rethinking the Racial Wealth Gap, 4 SOC. CURRENTS 199, 199–200 (2017) (describing
the targeting of minority homebuyers and students who borrow to fund mortgage or education
debt as predatory inclusion); Richard Rothstein, A Comment on Bank of
America/Countrywide’s Discriminatory Mortgage Lending and Its Implications for Racial
Segregation, ECON. POL’Y INST. (Jan. 23, 2012), https://www.epi.org/publication/bp335-boa-
countrywide-discriminatory-lending [https://perma.cc/28RB-BF6V] (describing the
Department of Justice’s settlement with Bank of America and concluding that “[t]he lending
industry seems to have systematically targeted African Americans and Hispanics for these
risky subprime loans”).
21. See generally Kathleen C. Engel & Patricia A. McCoy, From Credit Denial to
Predatory Lending: The Challenge of Sustaining Minority Homeownership, in SEGREGATION:
THE RISING COSTS FOR AMERICA 81 (James H. Carr & Nandinee K. Kutty eds., 2008).
22. Kevin P. Donovan & Emma Park, Perpetual Debt in the Silicon Savannah, BOS. REV.
(Sept. 20, 2019), http://bostonreview.net/class-inequality-global-justice/kevin-p-donovan-
emma-park-perpetual-debt-silicon-savannah [https://perma.cc/BPQ4-XMBZ] (exploring
fintech platforms in Kenya that offer access to credit “with speed and ease” to “millions of
Kenyans in need”); Wonga’s Woes Spell the End of the Payday-Loan Era, ECONOMIST (Aug
30, 2018), https://www.economist.com/britain/2018/08/30/wongas-woes-spell-the-end-of-
the-payday-loan-era [https://perma.cc/9QCH-XMUR] (describing the fall of the British
payday lender Wonga).
23. Donovan & Park, supra note 22. For other examples of fintech firms offering payday-
style loans, see CASHNETUSA, https://www.cashnetusa.com/ [https://perma.cc/W53B-4Y7F]
(last visited Oct. 6, 2019) and Sean Farrell, Where Did It All Go Wrong for Wonga?,
GUARDIAN (Aug. 7, 2018), https://www.theguardian.com/business/2018/aug/07/wonga-loans-
where-did-it-all-go-wrong [https://perma.cc/THR6-C2DL].
24. Donovan & Park, supra note 22.
剩余31页未读,继续阅读
资源评论
百态老人
- 粉丝: 1707
- 资源: 2万+
下载权益
C知道特权
VIP文章
课程特权
开通VIP
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功