Mitigating Legal Risks of Artificial

Intelligence: Best Practices for CounselFiduciary Expectations, Legal Challenges, Cross-Border Dealings, Managing Transactions

and Data Privacy

Today’s faculty features:

1pm Eastern | 12pm Central | 11am Mountain | 10am Pacific

The audio portion of the conference may be accessed via the telephone or by using your computer's

speakers. Please refer to the instructions emailed to registrants for additional information. If you

have any questions, please contact Customer Service at 1-800-926-7926 ext. 1.

WEDNESDAY, APRIL 25, 2018

Presenting a live 90-minute webinar with interactive Q&A

Robert W. (Bob) Kantner, Partner, Jones Day, Dallas

Michael W. Kelly, Partner, Squire Patton Boggs, San Francisco

Huu Nguyen, Partner, Squire Patton Boggs, New York

Dennis Garcia, Assistant General Counsel, Microsoft, Chicago

Tips for Optimal Quality

Sound Quality

If you are listening via your computer speakers, please note that the quality

of your sound will vary depending on the speed and quality of your internet

connection.

If the sound quality is not satisfactory, you may listen via the phone: dial

1-866-961-8499 and enter your PIN when prompted. Otherwise, please

send us a chat or e-mail sound@straffordpub.com immediately so we can address

the problem.

If you dialed in and have any difficulties during the call, press *0 for assistance.

Viewing Quality

To maximize your screen, press the F11 key on your keyboard. To exit full screen,

press the F11 key again.

FOR LIVE EVENT ONLY

Continuing Education Credits

In order for us to process your continuing education credit, you must confirm your

participation in this webinar by completing and submitting the Attendance

Affirmation/Evaluation after the webinar.

A link to the Attendance Affirmation/Evaluation will be in the thank you email

that you will receive immediately following the program.

For additional information about continuing education, call us at 1-800-926-7926

ext. 2.

FOR LIVE EVENT ONLY

Program Materials

If you have not printed the conference materials for this program, please

complete the following steps:

• Click on the ^ symbol next to “Conference Materials” in the middle of the left-

hand column on your screen.

• Click on the tab labeled “Handouts” that appears, and there you will see a

PDF of the slides for today's program.

• Double click on the PDF and a separate page will open.

• Print the slides by clicking on the printer icon.

FOR LIVE EVENT ONLY

Mitigating Legal Risks of Artificial

Intelligence: Best Practices for

Counsel

Fiduciary Expectations, Legal Challenges, Cross-Border Dealings,

Managing Transactions and Data Privacy

Dennis Garcia

Robert Kantner

Michael Kelly

Huu Nguyen

Scope of Talk

• Scope

• AI Primer – Demystifying AI

• AI tools in the legal industry and elsewhere

• Protecting AI data in the workplace

• AI’s impact on the workplace

• AI and transportation

• Ethical usage of AI

6

This Photo by Unknown Author is licensed

under CC BY-NC-ND

What is artificial intelligence and what are related legal challenges?

What is AI?

• Application of software that mimics human intelligence to perform tasks

normally performed by people

• Examples range from reading natural language such as English, predicting

outcomes of various problems, recognizing images, driving cars

Legal Challenges For In-House Counsels

• How to advise about AI in order for to meet fiduciary duties?

• Complying with established laws, such as anti-discrimination law when using

AI’s data and algorithms

• Liability when things go wrong

• Who owns algorithms and data for intelligence or should it be owned at all?

7

Artificial Intelligence

Examples

• Natural Language Processing

• Facial Recognition

• Autonomous Vehicles (“AV”)

• Expert systems

8

This Photo by Unknown Author is

licensed under CC BY-SA

This Photo by Unknown Author is

licensed under CC BY-NC

This Photo by Unknown Author is licensed

under CC BY

Brief Overview of AI for Non-Techies

Types of AI Technology

• Robotics

• Non-Physical Bots

• Natural Language Processing (NLP)

• Parsing, Translation, Generation

• Vision and image recognition

• Machine Learning

More above Machine Learning

• Supervised, Unsupervised, Semi-supervised and Reinforcement Learning

• Online vs offline Learning

• Symbolic, logic, statistical and neural methods

• Training data

9

This Photo by Unknown Author

is licensed under CC BY-NC-

SA

Big Tech Making Major AI Investments

10

AI Usage Growing Across All Industries

11

AI Examples Outside of Legal Industry

12

AI Examples Outside of Legal Industry

13

AI in the Legal Industry

14

AI Use Cases/Tools in Legal Industry

• Administrative Legal Support

• Do It Yourself (DIY) & Enhance Self-Help Resources

• Compliance: BrainSpace; NexLP; Neata Logic

• Contract Review/Due Diligence: eBrevia, Kira Systems; KM Standard, Ravn Systems,

LawGeex

• Discovery: Casepoint; Kcura; Modus Ediscovery

• Legal Research: Fast Case; Gavelytics; IBM Watson Debater; Lex Machina, Ross Intelligence

• Cybersecurity

15

Microsoft Legal Team’s Use of AI

16

Sales Team or Reseller

creates a sales quote

QUOTE DEAL

Run algorithm against

quote to identify

riskiest deals based on

data attributes in real-

time

Location

India.

Partner

X com

Sector

Commercial

Risk Score

5

China

Iceland

Robert Finn

Jason Hunt

Public Sector

Commercial

90

20

!

!Hungary

Brazil

Moneymart

PAT

Public Sector

Commercial

80

40

!

Risk Score

Geo

Partner

Customer

Discount and Dollar

Analytics monitor risks

and deal is approved if

controls are satisfied

Bank Reconciliation

SECURE BANK JaneCustomer

Debit Card 30.00

13.00 Credit JohnSmith

Debit Card

Paycheck

PaymentTransaction type

Payor Transaction type

XYZ Corp

Review identified

deals and conduct

control checks and

mitigate risks with

stakeholders

821

0 100

RISK-SCORE DEAL

Risky attributes detected:• Sales Geo

• Reseller

• Customer Vertical

• Deal Revenue

• Deal Discount

Microsoft Legal Team’s Use of AI

Ethical Issues for Lawyers

Model Rule 1.1: Competence

• Comment 8: “. . .[A] lawyer should keep abreast of changes in the law and its

practice, including the benefits and risks associated with relevant technology. . .”

• Comment 1 allows consultations with experts

Model Rule 1.16: Confidentiality:

• When AI is used to evaluate client information, reasonable measures to protect

confidentiality must be taken

• Lawyers must learn about and train on security measures in the software

Model Rule 2.1: Independent Judgment

• Does programming reflect any bias?

• Reliance on expert?

• No explanation?

18

Ethical Issues for Lawyers

Model Rule 5.3: Duty to Supervise Non-Lawyers

• Comment 3 to Rule: “Lawyers may use non-lawyers outside the firm to assist the

lawyer in rendering legal services to the client.”

• Non-lawyer assistance may include firms that rely on machine intelligence.

• Lawyer can share responsibility for supervising A.I. firm with an expert.

Model Rule 5.5: Unauthorized Practice of Law:

• Lawyer needs to understand AI

• Lawyer needs to review final work product

19

Protecting AI Data: Trade Secrets and Confidential Information

• State laws based on UTSA Sec. 1.4: "Trade secret" means information, including

a formula, pattern, compilation, program, device, method, technique, or

process, that: (i) derives independent economic value, actual or potential,

from not being generally known to, and not being readily ascertainable by

proper means by, other persons who can obtain economic value from its

disclosure or use, and (ii) is the subject of efforts that are reasonable under

the circumstances to maintain its secrecy.

• Defend Trade Secret Act of 2016

• What is the main form of AI trade secret?

• Algorithms and processes whether or not patentable

• Other AI Data such as neural networks, training data and sensor data

20

Protecting AI Data: Trade Secrets and Confidential Information

• “AI Data” means algorithms and processes, data, corpus, training sets,

parameters, neural network, rules, sensor data, etc. …

• Example is the Light Detection and Ranging (LiDAR) data that is the subject of

the Waymo v. Uber suit

• Issue

― Since most AI “knowledge” is stored as data or information, Trade

Secret protection is the ideal means for protecting the information

― Having confidentiality provisions is necessary (but not sufficient) to

keep information secret.

― Trade Secrets protection should survive the term of the

confidentiality agreement.

― Include Defend Trade Secrets Act Notice in NDAs

21

This Photo by Unknown Author is

licensed under CC BY

AI Dispute - Waymo LLC vs Uber Technologies, Inc.

• Waymo [LLC] sued Uber Technologies, Inc. (“Uber”), Ottomotto LLC, and Otto

Trucking LLC (together, “Ottomotto”) in the U.S. District Court for the

Northern District of California (“District Court”) alleging, inter alia, claims of

patent infringement and violations of federal and state trade secret laws.

Specifically, Waymo alleges that its former employee, Mr. Levandowski,

improperly downloaded thousands of documents related to Waymo’s driverless

vehicle technology and then left Waymo to found Ottomotto, which Uber

subsequently acquired. Waymo LLC v. Uber Techs., Inc (Fed. Cir. 2017); http://www.cafc.uscourts.gov/sites/default/files/opinions-orders/17-2235.Opinion.9-11-2017.1.PDF

• Waymo asked for $1.859 billion for Trade Secret 25 … and less for the

remaining trade secrets. Waymo LLC v. Uber Techs., Inc. (Waymo I), No. 17-cv-

00939-WHA (JSC), 2017 WL 2485382; https://www.documentcloud.org/documents/4059938-1778-

Waymo-s-Opposition-to-Defendants-Mtn-to.html

• February 2018, Uber agreed to pay Waymo $245M to settle the suit. https://www.reuters.com/article/us-alphabet-uber-trial/waymo-accepts-245-million-and-ubers-regret-to-settle-self-driving-

car-dispute-idUSKBN1FT2BA

22

Defining Confidential Information – Protecting Intellectual Property

• Exclusions from Confidential Information. Except as required by applicable

federal, state, or local law or regulation (“Law”), and except for [Supplier’s]

[or Customer’s] AI Data which will remain confidential, the term

“Confidential Information” as used in this Agreement shall not include

information that falls within the following exceptions (“Confidentiality

Exceptions”): (a) at the time of disclosure is, or thereafter becomes,

generally available to and known by the public other than as a result of,

directly or indirectly, any act or omission by the Recipient/Customer or any of

its Representatives; …

• Issue

• Should publicly available data be excluded and how?

• https://www.nytimes.com/2018/02/05/technology/uber-waymo-trial.html

• Uber’s lawyers argued in its opening statement that their technology was

similar to Waymo’s because it was based on publicly available data

• Takeaway: Protect AI Data as Confidential Information in addition to as

Trade Secrets23

AI in the Workplace - Fair Credit Reporting Act

- FCRA also applies to background screening companies and may apply to data

brokers.

- FTC entering into consent decrees with several data brokers which

advertise their consumer profiles for employment screening purposes

- FCRA requirements:

- Consumer Reporting Agencies implement reasonable procedures to ensure

maximum accuracy of reports

- Consumer access to review and correct information

- Notice if an adverse decision is made based on a consumer report

- Federal Trade Commission warning: “companies should be mindful of the law when

using big data analytics to make FCRA-covered eligibility determinations.”

- How does a consumer review an AI’s “file” or get to “dispute incomplete or

inaccurate information”?

- How do you notify when an AI does not select a candidate for an interview?

24

Employment Discrimination by Robot?

- Use of AI algorithm and big data to make employment decisions

- What is the algorithm basing the decisions on?

- Disparate Impact versus Disparate Treatment

- Age, Race, Gender, Disabilities, etc.

- Algorithms cannot directly disfavor protected groups

- Some features of algorithms may have a disproportionately adverse impact

on a protected class

- An AI “learns” that best employees come from suburbs which tend

to have largely white populations

- Understand the difference between “correlation” and “causation”

25

AI and Workplace Privacy – Workplace Monitoring

• Why Monitor?

• Recording activities for training an AI

• Unknown monitoring by vendors for their own AIs

• Cybersecurity

• Protecting company assets and information

• Mobile devices, IoT devices and tracking.

• When can employers monitor GPS and geo track employees?

• Some states require consent of owner of device or vehicle to track

• some states permit employees to sue under common law invasion of privacy or

other state privacy laws

• Medical privacy issues for tracking fitness wearables (e.g., in relation to

employee wellness programs)

26

AI and Workplace Privacy – Workplace Monitoring

• Employees privacy protection in the workplace:

• Computer Fraud and Abuse Act

― Lee v. PMSI, Inc., 2011 (breaching an employer's acceptable use

policy was not "unauthorized access" under the act and, therefore,

did not violate the CFAA)

• State privacy torts and constitutional privacy rights

• Recommendation

• Use written, disseminated policies to notify employees they will be

monitored

• Narrow employees’ expectation of privacy in the workplace

• Include an acknowledgement by employees that they will be minitored

27

Artificial Intelligence and Workplace Torts

Despite fears of displacement, automation in industry and other

sectors is predicted to lead to:

• a safer and more efficient workplace environment

• a decrease in worker injuries

• workers compensation savings for employers

28

Artificial Intelligence and Workplace Torts

29

http://www.riken.jp/en/pr/press/2015/20150223_2/

Artificial Intelligence and Workplace Torts

Risks Still Exist!

While overall workplace safety should be improved, introducing robots into the work

environment with humans does create a risk of injury from the robots themselves—with

the fault potentially resting with the employer (e.g., failure to implement safety

protocols), the employee (e.g., misuse or failure to observe safety protocols), the

manufacturer (e.g., for defective design), or other third parties.

30

Artificial Intelligence and Workplace Torts

Product Liability Claims

• Under a traditional framework, when someone is injured using a product, product

liability law determines when and to whom liability attaches

• Product liability is a creature of state common and statutory law:

• Negligence (duty, breach, causation, damages)

• Breach of warranty (merchantability or fitness for a particular purpose)

• Strict Liability (RST 2d Torts § 402A )

― Design defect

― Manufacturing defect

― Failure to warn

31

AI in Transportation

32

Google Self Driving Car Project

Est. 2009

33

Waymo Self Driving Car Project

2018

34

Autonomous Trucks

35

Autonomous planes

36

Autonomous Boats

37

Autonomous Trains

38

Autonomous Cars – South Korea

39

May 9, 2017

Autonomous Cars – Japan

40

February 15, 2017

June 6, 2017

September 8, 2016

Autonomous Cars – China

41

July 5, 2017

Regulatory Risks – Technical Challenges

• Neural networks are so called because they mimic – approximately – the

structure of the brain. They are composed of a large number of processing

nodes that, like individual neurons, are capable of only very simple

computations but are connected to each other in dense networks.

• In a process referred to as "deep learning," training data is fed to a network's

input nodes, which modify it and feed it to other nodes, which modify it and

feed it to still other nodes, and so on. The values stored in the network's

output nodes are then correlated with the classification category that the

network is trying to learn – such as the objects in an image, or the topic of an

essay.

• Over the course of the network's training, the operations performed by the

individual nodes are continuously modified to yield consistently good results

across the whole set of training examples. By the end of the process, the

computer scientists who programmed the network often have no idea what the

nodes' settings are. Even if they do, it can be very hard to translate that low-

level information back into an intelligible description of the systern's decision-

making process.

42

Regulatory Risk – Technical Challenges

D

O

G

D

2

3

1

2

3

4

5

6

7

8

INPUTNODES

PERFORMANCENODES

OUTPUTNODES

43

Regulatory Risk – Technical Challenges

D

O

G

D

O

X

2

3

4

5

9

8

7

6

INPUTNODES

PERFORMANCENODES

OUTPUTNODES

44

Regulatory Risk – Technical Challenges

D

O

G

D

O

G

3

4

5

6

10

9

8

7

INPUTNODES

PERFORMANCENODES

OUTPUTNODES

45

Regulatory Risk – Technical Challenges

D

O

G

D

O

G

6

5

4

3

7

8

9

10

INPUTNODES

PERFORMANCENODES

OUTPUTNODES

46

Liability Risks – Proof of Design Defect

• In re Toyota Motor Corp. Unintended Acceleration Products Liability Litigation, 978 F.

Supp. 2d 1053 (C.D. Cal. 2013) (applying Georgia law)

• Executor of driver’s estate sued automobile manufacturer to recover damages she

sustained when her vehicle unexpectedly accelerated allegedly without her depressing

the gas pedal.

• One issue was whether plaintiff’s expert would be allowed to testify that the car had a

design defect even though he could not identify with certainty a specific software bug

(or other specific cause) that could have opened the throttle from its idle position.

• The expert did assert there were errors in the source code, including:

• Inadequate operating system;

• Substandard ECM software architecture;

• Negligently designed watchdog supervisor software;

• Untestable, unduly complex nature of the “spaghetti” code;

• Task X could disable the fail-safes and cause unaccelerated acceleration;

• An unidentified software bug could cause partial task death of Task X.

• (Task X calculates throttle angle, monitors for system failure and enters fail

safe modes.

47

Proof of Design Defect

• The court determined that Georgia law did not require identification of a specific

defect.

• To determine liability for alleged design defects, Georgia applies a risk-utility test.

This test incorporates the concept of reasonableness, i.e. whether the manufacturer

acted reasonably in choosing a particular product design, given the probability and

seriousness of the risk posed by the design, the usefulness of the product in that

condition, and the burden on the manufacturer to take the necessary steps to eliminate

the risk.

• All the plaintiff must show is that the device did not operate as intended and this was

the proximate cause of his/her injuries.

• Design defects may be proven through circumstantial evidence. This is especially true

when:

• The alleged defect destroys the evidence necessary to prove the defect; or

• When the evidence is otherwise unavailable through no fault of the plaintiff.

• Here, plaintiff’s expert testified that the car’s software does not record software

failures. That’s good enough.

• Motion to strike expert report denied. Motion for summary judgment denied.

• Case settled.

48

Technical Challenges to Regulation

House of Commons Science and Technology Committee 2016 Report on Robotics and Artificial

Intelligence explained the need to ensure AI operates as intended

According to the Association for the Advancement of Artificial Intelligence (Menlo Park,

CA):

It is critical that one should be able to prove, test, measure and validate the

reliability, performance, safety and ethical compliance–both logically and

statistically/probabilistically – of such robotics and artificial intelligence systems

before they are deployed.

Similarly, Professor Stephen Muggleton, Professor of Machine Learning at Imperial College,

London, saw a pressing need:

To ensure that we can develop a methodology by which testing can be done and the

systems can be retrained, if they are machine learning systems, by identifying

precisely where the element of failure was.

But the verification and validation of autonomous systems is “extremely challenging”

since they are increasingly designed to learn, adapt and self-improve during their

deployment. Traditional methods of software verification cannot extend to these

situations.

49

Technical Challenges to Regulation

The House of Commons Report posed the challenge:

It is currently rare for AI systems to be set up to provide a reason for reaching a particular

decision. For example, when Google DeepMind’s AlphaGo played Lee Sedol in March 2016,

the machine was able to beat its human opponent in one match by playing a highly

unusual move that prompted match commentators to assume that AlphaGo had

malfunctioned. AlphaGo cannot express why it made this move and, at present, humans

cannot fully understand or unpick its rationale. As Dr. Owen Cotton-Barratt from the

Future of Humanity Institute reflected, we do not “really know how the machine was

better than the best human Go player.”

. . . .

Part of the problem [is] that researchers’ efforts [have] previously been focused on

achieving slightly better performance on well-defined problems, such as the classification

of images or the translation of text while the “interpretation of the algorithms that were

produced to achieve those goals had been left as a secondary goal.”

50

Possible Regulation – An Explanation

• Harvard University Berkman Klein Center Working Group on Explanation and the Law

says:

• A.I. / A.I. devices should give the reasons or justifications for a particular

outcome, but not a description of the decision-making procedures.

• Also, we need to know whether changing a factor would change the

decision.

• This explanation should be given whenever a human (or corporation)

would have to give an explanation.

• But what about the weighing of factors? Judgment?

51

Google Researchers Say They’re Learning How Machines Learn

52

March 7, 2018

Google Research

53

March 7, 2018

Inside a neural network, each neuron works to identify a particular characteristic that might show up in a photo, like a line that curves from right to left at a certain angle or several lines that merge to form a larger shape. Google wants to provides tools that show what each neuron in trying to identify, which ones are successful and how their efforts combine to determine what is actually in the photo – perhaps a dog or a tuxedo or a bird.

Google Research

54

March 7, 2018The kind of technology Google is discussing

could also help identify why a neural network is prone to mistakes and, in some cases, explain how it learned this behavior; Mr. Olah said. Other researchers, including Mr. Clune, believe they can also help minimize the threat of “adversarial examples” – where someone can potentially fool neural networks by, say, doctoring an image.

Ethical Usage of AI

55

Source: “The Future Computed: Artificial Intelligence and its role in society,” January 17, 2018.https://blogs.microsoft.com/blog/2018/01/17/future-computed-artificial-intelligence-role-society/

56

Ethical and Risk Mitigation Issues

Real ethical issues in “no good outcome” situation

Development of AI:

• Explanation

• Document Retention

Ethical Issues for Lawyers

• Competence

• Witness?

• Compliance

57

Practical Takeaways

58

About the Speakers

Huu Nguyen

Squire Patton Boggs, New York, huu.nguyen@squirepb.com

Mr. Nguyen is a deal lawyer, focusing his practice on commercial and corporate transactions in the

technology space. Prior to being an attorney, Huu was an AI programmer, and research scientist.

Mike Kelly

Squire Patton Boggs, San Francisco, michael.kelly@squirepb.com

Mr. Kelly has experience in employment litigation, counseling, collective bargaining and arbitration. His

practice includes state and federal employment litigation regarding wage and hour issues, age and disability

discrimination, sexual harassment and retaliation.

Robert Kantner

Jones Day, Dallas, rwkantner@jonesday.com

Bob Kantner is an experienced trial lawyer who handles intellectual property and business cases. Bob has

gone to trial 36 times (half IP and half business), serving as lead trial counsel in 28 of those trials; his clients

have prevailed outright in 75 percent of the trials. Bob organized a working group at Jones Day focused on

artificial intelligence and robotics and serves on another working group focused on autonomous vehicles. He

has written a white paper on legal issues related to the development of autonomous and connected cars and

has written and spoken about artificial intelligence numerous times, most recently on the subject of ethical

issues for lawyers using artificial intelligence.

Dennis Garcia

Microsoft Corporation, Chicago, dennisga@microsoft.com and @DennisCGarcia on Twitter

Mr. Garcia practices at the intersection of law, technology and business and provides a wide range of legal

support to Microsoft’s US Sales, Marketing & Operations teams.

59