A
ABA. See American Bar Association
ABS. See alternative business structures
accountants, 34
acknowledgements, xii
adoption of AI, 95–115
and change management, 60, 63, 66, 107, 111-112
and engagement, 63, 66
as enterprise-wide transformation, i, 202–204
barriers to, 2, 96-97, 108, 111, 113–115
user adoption, 53, 125–127, 133–134, 139
agents and agentic AI,
case study, 186
definition, 3–5
emergence of, 31–36
in law firm workflows, 165–166, 180
using data, 80, 85–93, 147
AI benchmarking, 128
AI committee, 62–66, 76, 77, 121, 139, 153, 156
AI engineers. See legal engineers
AI-first law firms, 9, 13–14
AI lawyers, 105–106
AI literacy. See Literacy
AI policy, 6, 45, 63, 131, 152–156
alignment (of knowledge and innovation), 61
alternative business structures (ABS), 12–13
alternative fee arrangements (AFAs), 9–10, 38–39, 41
Altman, Sam, 25
American Bar Association (ABA), 144–145
analytics. See data
Anthropic, 13, 25
Cowork. See Cowork
Microsoft integrations, 175–176
Model Context Protocol (MCP). See Model Context Protocol (MCP) 67
models, 70
skills, 130,
See also Claude (AI Assistant)
API (application programming interface), 88
apprenticeship, 103
Artificial Lawyer (publication), 9, 105, 175
“Attention Is All You Need” (Vaswani et al.), 23
attention mechanism, 23–24
audit and audit trails, 17, 187, 192, 204
authors (about), 219
automation,
document, 28
effects on training, 95, 102
exposure to, 48–49, 54
workflow, 73, 120–121
B
Bamford, Catherine, 28
BamLegal, 28
Bar Council, 106
Bar Professional Training Course, 101
Base44, 159
Benamram, Oz (author), 15, 194
benchmarking
AI. See AI benchmarking
outside counsel, 2, 17, 179
BERT (Bidirectional Encoder Representations from Transformers), 24–25
bias, 86, 123, 142, 144, 146
BigLaw, 14, 39[AC2]
billable hour, 1–2, 47–48, 190
business mode, 143
death of, 9, 16
engagement, 112–113
Jevons paradox, the, 151–152
pricing, 38–39, 41,
rewarding, 11, 53, 55, 168
targets, 11–13, 108–109, 115, 170
billing systems, 81
Bloom, Benjamin, 99
build vs buy, 66, 132–133, 163, 168
C
career development, 104–106, 204
case law databases, 19, 23, 30–31, 87
CCPA (California Consumer Privacy Act), 147
change management, 60, 63–66, 111–112, 189
ChatGPT, 24–25, 63, 74, 147, 149, 161, 203
See also OpenAI
Chief Executive Officer, 25
Chief Innovation Officer, 65, 74
Chief Operating Officer, 25
classification (AI function), 30, 36
Claude (AI assistant), 25, 63, 67, 74, 160–161, 175–176
Claude Code, 133, 163, 165–166
Claude Cowork, xi, 160, 165–166
Claude for Word, 175
See also Anthropic
Clifford Chance LLP, 75
Clio, 107–108
CLOC (Corporate Legal Operations Consortium), 195
Codex, 133
Cohen, Joe, 105
competence (professional),
professional, 99, 144–146, 152, 155
technology, 72, 144–146
competitive advantage, 92, 110, 165, 175, 182, 191–192
compliance, 21, 58, 64–65
as a client service, 68
risk, 71,
standards for technology, 137–138
with codes of conduct, 145
confidentiality, 88, 92, 146–148, 150, 152, 155, 196
consultants, role of, 196–197
context window, 26, 29, 35
contracts, drafting and review of,
drafting, 19
extraction from, 30
playbooks, 7, 167
contribution hours. See innovation hours
Corporate Legal Operations Consortium. See CLOC
Cowork. See Claude Cowork
Curphey, Adam (author), vii, 194, 69, 120
customer success, 117, 133
D
dashboards, 6, 66, 80–82, 186, 195
data
analytics, 66, 82, 102, 178
as work product, 79–93
dashboards, see dashboards
data analyst, 34, 70, 91
data architects, 70–71, 93
data engineers, 70
data governance, 44, 93, 131, 136, 141–156, 163
data infrastructure, 87–88
data protection, 146–148, 152, 205
data quality, 92
data science, 66, 70–71, 92, 102
hygiene, 42, 29, 54
metadata, 83–84, 87, 92–93
proprietary datasets, 87, 165, 175
structured data, 80, 91, 126
synthetic data, 163
training data, 26, 86, 89–90
DeepJudge, 29, 163
deep learning, 23
delta, 10, 50, 55
deployment, 5–7, 58–66, 77, 120–121, 138
agents, 33
case study, 45
cloud, 148
early days of, 29
process mapping, 73
strategy, 38
differentiation, 46, 182, 204
discovery. 29,
See also eDiscovery
document automation, 28
document management systems (DMS), 67, 81, 118, 132, 136, 139, 183
drafting,
training, 97–99, 114, 129
with AI, 28, 30–31, 36, 81–82, 92, 129, 205
“drinking your own champagne”, x–xi
See also “eating your own dog food”
due diligence,
agent, 15
as a feature of AI tools, 30–32
before generative AI, 67
case study, 3–5, 10, 41, 186
charging, 43, 51
getting started, 205
platform selection, 129, 131
training, 96–97, 105
E
early adopters, 60, 64, 113–114
“eating your own dog food”, x–xi
eDiscovery, 19, 31, 40
charging for, 43, 160–161
point solutions, 108, 118, 129
training, 105
effective fee arrangements (EFAs), 39
encryption, 137
engagement, 95–115
engagement letter, 82, 93
England and Wales, 100–102, 145, 181, 197
enterprise search, 29, 67, 132
enterprise tools, 90, 107, 135, 160, 163–166, 176
ethical walls, 89, 136
ethics, 143
bias. See bias
environmental, 146
EU AI Act, 146
evaluation (of AI tools), 61, 123–125, 127–128, 137–138, 168
Everlaw (eDiscovery), 129
Excel (Microsoft), 175
extraction (AI function), 30, 36
F
feedback and feedback loops,
client, 89, 184
collection, 66, 125–126
data, 83
loop, 194
training, 98–99, 114–115
fees, 9–10, 38–39, 40–42
fine-tuning, 26, 89
fixed fee, 47, 55
foundation models, 68, 130, 175
frontier models, 118, 158, 176
functional categories of AI, 36
G
Garfield.Law, 14
GDPR (General Data Protection Regulation), 147
Gemini (Google), 25, 67, 166
general counsel, 62–63, 68, 152–153, 178, 196
generative AI, 3–5, 20–21, 23, 25–32, 35–36, 69
Georgia Tech, 98
Google, 13, 23–25, 67, 147, 183–184
governance,
AI, 6, 16, 34
AI committee, 63, 76
data, 44, 86, 89, 92–93, 129, 131, 141–156
information, 68, 71–72
GPT, 16, 24
See also ChatGPT
H
hallucinations, 26–27, 29, 35, 72, 86, 141
Harbor Global, 197
Harvey (AI), 29, 67, 103, 108, 129, 131, 163–164
human review, xi, 86, 92, 110, 120, 191, 204
human skills, 188–189, 192
I
ILTA (International Legal Technology Association), 195
in-house teams and legal departments,
AI committee, 77
benchmarking outside counsel, 16–17
billable hour, 11
buying from legal vendors, 14
change management, 112
data, 88, 90
experimenting with AI, 158
Jevons paradox, the, 152
performing new tasks, 7–9, 43–45, 51, 111–112
shadow AI, 174–192
strategy, 1–3, 8, 37–39
SQE, 101
value, 39
where AI lives in, 64
innovation
innovation hours, 11, 53, 55
innovation lawyers. See AI lawyers
innovation officer / chief innovation officer, 65, 74
innovation team,
composition, 59–60, 65
driving adoption, 95
empowering lawyers, 73–74, 76
partners leading, 106
shadow AI, 157
working with lawyers, 127, 151
working with vendors, 117, 122–124, 129, 134, 192
working with risk, 153, 156
workload, 25
intellectual property, 89, 184
International Legal Technology Association. See ILTA
investment banking, 81, 102
ISO 27001, 137
IT (teams), 60, 62–63, 65, 121, 133, 135
J
Jevons paradox, the, 51, 151–152
Joyner, Professor David, 98
junior lawyers, 3, 8, 75, 105, 115, 184
training, 96–99
K
Kira (Kira Systems), 67
knowledge engineers. See legal engineers
knowledge hours. See innovation hours
knowledge management (KM), 5, 52, 58–60, 65, 71, 79, 194–195
Know Your Customer (KYC), 83
Kraft Kennedy, 197
L
LaaS. See law-as-a-service.[AC3]
large language models (LLMs), xi, 2, 23–27, 30, 35, 142
law-as-a-service (LaaS), 15
LawFairy, 13
law firms
alternative business structures, 12–13
billing models, 1, 7, 9–13, 38–48
law firm–client relationship, 37–55
leverage / pyramid model, 38, 62, 75–76, 96, 105, 131, 174
partnership model, 13, 50, 62
smaller firms and sole practitioners, 7, 197
law schools, 100–103
Law Society, 197
LawtechUK, 197
lawyer-built solutions, 157–171
Legal Cheek, 102, 109
legal engineers, 19, 66, 68–69, 71, 91, 93, 187
building, 175
legal operations, 3, 44, 90, 195
LegalQuants, 68–70, 196
Legaltech Hub, 119, 196
Legal Services Act 2007, 12
Legora, 67, 103, 108, 129, 131, 163–164
leverage, 38, 62, 75–76, 96, 105, 131, 174
Levy, Colin, 194
lifelong learning, 113–114
LinkedIn, 117, 119, 134, 194,
literacy (technical and AI), 34, 90–91, 101, 106
listening tour, 120
litigation,
AI-first law firms, 14
data, 86
eDiscovery. See eDiscovery.
prediction, 19, 30
privilege, 148–149
witness statements. See witness statement
LLaMA (Meta), 25
LLM. See large language models
lock-step, 104
Lovable, 133
LPC (Legal Practice Course), 101
Luminance, 67
M
M&A (mergers and acquisitions), 15, 41, 98, 107, 127
Macfarlanes, 45
machine learning, 22–23, 70, 136
managed services organization (MSO), 13
Matrix, the, 80, 82
mentoring, 98–100
Meta, 13, 25
metadata, 83–84, 87, 92–93
Microsoft, 13, 175–176, 180, 183–184, 194
Microsoft Copilot, 131, 180, 183
Microsoft Outlook, 79, 81
Microsoft Word, 184
moats, 182, 191
Model Context Protocol (MCP), 33, 36, 87–88, 92, 130, 182, 190
multimodel (AI architecture), 117, 128
Munir v Secretary of State for the Home Department, 149
N
natural language interfaces, 2–3, 21, 33, 42, 81–82
natural language processing (NLP), 21
neural networks, 23–24, 75
New York State Bar Association, 197
no-code tools, 157, 195
Northwestern Pritzker School of Law, 102
O
OpenAI, 13, 24–25, 70, 175
See also ChatGPT
Orbital, 129
Orbital Copilot, 131
Orchestration, 166, 170
Outlook. See Microsoft Outlook
P
panel reviews (client), 41, 43, 88, 152, 181
paralegals, 13, 180
partners (law firm),
AI committee, 64–65, 121–124
building solutions, 159, 161
earnings, 9
incentives, 109–110, 113
leading innovation, 106
partnership model. See partnership model
partnership track, 104–105, 170
reviewing data, 82
selecting tools, 139
training, 98–99, 115, 151, 156
understanding AI, 106–107
working with innovation, 61, 75
partnership model, 13, 50, 62
Pasternak, Rebecca (author), 194
pilots and pilot programmes, 52, 192, 198–199, 205
design, 125–126
procurement, 133, 139
persona, 69, 72, 74
playbooks, 7, 15, 44, 83, 109
point solutions, 108, 118, 140, 166
policy. See AI policy
practice solutions team, 66
precedent banks, 81, 88
predictive
AI function, 30, 36
analytics, 70
coding, 22
pricing, 9–13, 16, 37–49, 54, 113, 160
data, 82–84, 113, 160
prioritization, 49, 64, 120–122
privacy, 68, 85, 141–156
privilege, 143–144, 148–150, 152, 155
probabilistic outputs, 26, 28, 142
process mapping, 73, 120, 139
procurement, 121–122, 130, 133–136, 198
product managers, 66
professional conduct, rules of, 144–146
profit and loss account, 110
project managers, 34
prompt engineering and prompt engineers, 69, 72
prompts and prompting, 3–5, 26–27, 29, 32, 69, 72, 192
proof of concept (POC), 202, 205
pyramid model. See leverage
Q
Qualifying Work Experience (QWE), 101
R
RAG (retrieval-augmented generation), 28–29, 35
real estate, 4, 15, 47, 129, 131
realization, 42, 83, 105
reasoning models, 21, 30, 36
recovery, 83
Relativity (eDiscovery), 129
request for proposal (RFP), 69, 74
retrieval, 28–29, 32, 35, 90
return on investment (ROI), 109–110, 202
revenue, 10, 37, 44, 48–54, 109–110, 160–161, 167–168
RFP. See request for proposal
Riehl, Damien, 13
risk, 141–156
ROI. See return on investment
Roscoe, Jasper & Mills LLP (RJM) (fictional firm), 4, 10, 82, 186
S
SaaS. See software-as-a-service
sandbox environments, 74, 135, 163, 170
security, 136–138, 162–165, 170, 183–184, 188
semantic search, 24
senior lawyers, 3, 8, 101, 106, 115
See also partners (law firm)
service design, 73
shadow AI, 57, 63, 76, 153–154, 157–171
Shoosmiths, 8
SKILLS (Strategic Knowledge & Innovation Legal Leaders’ Summit), 5–6, 119, 145, 150, 195
skills (AI), 174
skills (capability), 60, 68, 72–73, 98, 103–104, 188–189, 192
See also human skills
SOC 2 Type II, 137
soft skills. See human skills
software-as-a-service (SaaS), 174–175, 191
Solicitors Qualifying Examination (SQE), 101–102
Solicitors Regulation Authority (SRA), 14, 143–145, 149
SRA. See Solicitors Regulation Authority (SRA)
stochastic, 27
strategy, 37–55
structured data, 80, 91, 126
Susskind, Richard, i, 42
synthetic data, 163
T
T-shaped lawyer / persona, 69
TAR (technology-assisted review), 22–23, 170
team building, 57–77
time recording systems, 80–81
tokens and tokenization, 26–27, 30
Tomorrow’s Lawyers (Susskind), i
tool selection, 117–140
trainees. See Junior Lawyers
training, 95–115
transformation, i, 22, 53, 174, 191, 202
function 66
triangle of transformation, 176–177
transformer architecture, 23–24, 35
transparency, 11–12, 16, 184, 190, 192, 207
Two-Face (Harvey Dent), 29
U
unauthorized practice of law (UPL), 13
United States, 100–102, 144, 181, 197
United States v. Heppner, 143, 149
use cases, 121–122, 129, 131–132, 153, 164
V
value-based pricing, 10, 54, 113, 160
Vaswani, Ashish, 23
vendors
in firm-client-vendor triangle, 174–192
procurement of, 121–122, 130, 133–136
selection and evaluation, 117–140
update problem, 134–136
vibe coding, 33, 154, 158–165, 169
Vincent, 107
vLex, 107
W
Warner v. Gilbarco, Inc., 143, 149
Waters, Tara, 105
witness statement, 30, 106
witness testimony. See witness statement
Wix, 159
workflow, 3, 13
AI, 43–44, 46–47, 148, 171, 183–185
agentic, 15, 34, 91–92, 129
as intellectual property, 89
automation, 49, 73, 120
building, 69–70, 158–159, 205
client, 177, 188
data, 83
legal, 32, 52–53, 67, 86
procurement, 133