Sie sind auf Seite 1von 30

1

AI for Legal Services


April 12, 2017
2

Agenda

AI defined
Hype versus reality
AI in the legal market today
Outlook for legal AI
Gaining benefits of legal AI
If cost reduction is goal why AI?
Conclusions

Proprietary Not for Circulation


3

AI DEFINED
4

AI Includes Multiple Technologies

Credit: Michael Mills, Neota Logic, 2015,


Proprietary Not forArtificial
CirculationIntelligence in the Law: The State of the Play 2016, Part 1
5

AI HYPE
6

Media Hypes Legal AI

Proprietary Not for Circulation


7

AI REALITY
8

But Media Also Questions AI

Proprietary Not for Circulation


9

Study Finds Limited AI Impact on Lawyers

Journal article by economists on the potential employment impact of


legal AI
Multiple legal market interviews
Analyzed e-billing data

Lawyer job loss projection:

13% over 5 years is the maximum impact

Assumes nothing changes, which they say is unlikely

Source: Dana Remus and Frank Levy, Can Robots be Lawyers? (SSRN, 30 Dec 2015)

Proprietary Not for Circulation


10

AI IN LEGAL TODAY

Proprietary Not for Circulation


11

Current Applications of Legal AI In-House

Type of Application Likelihood and Use Case


eDiscovery doc review Proven and deployed for doc
review

Contract analytics Likely for companies with high


transaction or contract volumes
that need to conduct due diligence

Interactive legal advice Likely via licensed or free use of


firm systems
Internal development:
questionable

Legal research Too early to know


Just being deployed
Lex Machina is a good example

Proprietary Not for Circulation


12

Current Legal AI Use is Narrow

Ross Intelligence (Watson) for Bankruptcy Law


BakerHostetler, Latham, a few other firms

Kira Systems and RAVN (machine learning)


Multiple law firms for due diligence
Two firms for contract management, perhaps more (BLP +
Linklaters)

Expert Systems
Akerman, Cadwalader, Foley, Hall & Wilcox (UK), Husch Blackwell,
Littler, Norton Rose.

Law departments??

Proprietary Not for Circulation


13

Two Contract Challenges To Consider for AI

Two new accounting standards may be costly to comply with


Revenue recognition
Bringing leases onto balance sheet

Both require
Finding contracts
Systematically extracting relevant terms
Analyzing those terms
Taking action to comply based on findings

Where are companies on this? What is role of General Counsel?

Proprietary Not for Circulation


14

AI Reality Summary: Augmented, not


Artificial

Source: Ryan McLead, Aha! A.H.I!, 3 Geeks and a Law Blog (very popular legal tech blog), Dec 2015

Proprietary Not for Circulation


15

AI Outlook in Legal:

Lawyers Adopt New Tech Slowly


16

Legal Tech Uptake is Slow. Why is AI


Different?

Manageme
Core Tech Practice Tech
nt
PCs Excel E-Billing
Internet access Instant messaging Enterprise Legal
Firm websites Document Management
Social media assembly (ELM)
Security Contract analytics Extranets: more
systems Document talk than action
authoring tools Metrics +
analytics

Proprietary Not for Circulation


17

The Three-Year Legal AI Horizon

No reason to expect AI path will differ from other tech


The legal market moves slowly
Business cases take a long time to prove
Law departments change at about the same rate as law firms
Late adopters almost never punished

Incremental change. No revolution

Proprietary Not for Circulation


18

Gaining the Benefits of AI


19

Process to Evaluate AI Potential Savings

1. Identify candidate work areas


Analyze all legal + compliance work
Determine low- to mid-level complexity tasks that recur regularly
2. Determine potential ROI
Measure current unit costs
Determine AI cost (license fee, deployment, etc)
Estimate future unit costs with AI
3. Pilot to confirm or reject ROI analysis
4. Ensure adoption if pilot succeeds
5. Ongoing metrics, evaluation, and adjustment

Proprietary Not for Circulation


20

Process to Evaluate AI Technology

Develop costs + outcomes metrics for current work approaches

Empirically test AI solutions


Cannot evaluate the back-end even the scientists argue
Focus on empirical testing
Requires designing appropriate trials with before + after metrics

Questions to answer
Who does the evaluation?
Who owns the decision?
When do you look at next generation of tech?

Proprietary Not for Circulation


21

If Goal is Reduce Legal Spend with


Tech

Does AI Offer the Biggest Savings?


22

How Much Existing Tech Have You Adopted?

Competence on basics (Word, Excel, PowerPoint, PDF)


Practice technologies
Document authoring
Deal management
IP management
Knowledge management
Legal project management

Proprietary Not for Circulation


23

If New is Good, Is AI the Right Focus?

Emergin
AI
g Tech

Big Data

Virtual
Reality

Bots
AI
Blockchain

Proprietary Not for Circulation


24

Emerging Tech Overview

Blockchain
Tech underlying Bitcoin (trustless ledger)
Smart contracts, faster transactions, govern the Internet of Things

Virtual Reality
Immersive experience
Simulates being there

Bots
Chat with an intelligent agent (text today, voice in future)
May replace mobile apps with texting

Big Data
Collect and analyze huge quantities of data
Changing retail, Internet of Things, manufacturing, urban analysis,
other markets
Proprietary Not for Circulation
25

Emerging Tech

Virtual
Blockchain Bots Big Data
Reality
Threa

Big shift in Hard to see Hard to see Competitors


financial now now bots use
t

markets + are just an AI effectively


commercial interface first
contracting
Opportuni

Cost savings Improve Cost savings


and better internal + and better Preventive
ty

customer external customer law


service comms service

Proprietary Not for Circulation


26

Emerging Tech with Real Potential

BLOCKCHAIN BIG DATA


Automate contracts Preventive law the biggest
opportunity. Do Less Legal
Reduce need for lawyer time
Work
Reduce friction internally +
Identify what gives rise to
externally
legal + compliance costs
Still uncertain
Then take steps to change
underlying corporate behavior
Who owns this responsibility?

Proprietary Not for Circulation


27

If Goal is Reduce Legal Spend

What Should You Look at Beyond


Tech?
Proprietary Not for Circulation
28

Non-Tech Approaches to Reducing Legal


Spend

Knowledge management
Legal project management
Process improvement
Train lawyers to use effectively the tech they already have
Do less law practice prevention

Proprietary Not for Circulation


29

Conclusion
30

Conclusion
Presumably the interest in AI is to reduce cost and exposure

But AI may be a solution in search of a problem

Short term, its impact on legal cost likely will be limited

So from a buying perspective


Rank AI with other potential answers to the problem
Pursue those with highest likelihood of success

If AI per se is the goal, dont rely on lawyers alone to figure out how
best to use it

Proprietary Not for Circulation

Das könnte Ihnen auch gefallen