Chapter 5 · Artificial intelligence

Artificial intelligence:
usage and ethical framework

Artificial intelligence is the topic that generates the most expectation in the sector. This chapter separates two questions that should not be confused: how AI is used today in legal work, and with what ethical safeguards it is being adopted.

The idea in one sentence

AI is present in most participating organizations. Daily use of at least one tool is a majority practice among respondents to the AI tools block, although frequency varies substantially by task. Risk assessment practices also vary.

How many organizations use AI (and an important clarification)

Around 7 out of 10 organizations report having implemented some AI tool, with no clear difference between audiences.

Imputation adds 9 firms and 8 departments with a blank adoption answer but completed AI tool items. With imputation: 53/72 firms (73.6%) and 25/36 departments (69.4%). Explicit answers only: 44/63 (69.8%) and 17/28 (60.7%). Both numerator and denominator change; three departments still have no valid adoption answer.

How AI is used today

Use of AI tools (Law firms)
Legal research
4.15
Translation
3.85
Document drafting
3.83
Document review
3.79
Document summarization
3.71
Contract analysis
3.42
Legislative change monitoring
3.38
Drafting of minutes
2.92
Legal due diligence analysis
2.74
Case management
2.40
Predictive case analysis
2.17
Use of AI tools (Legal departments)
Document drafting
3.96
Translation
3.96
Document review
3.92
Document summarization
3.88
Contract analysis
3.56
Drafting of minutes
3.46
Legal research
3.38
Legislative change monitoring
3.17
Legal due diligence analysis
2.67
Case management
2.52
Predictive case analysis
2.43

Usage scale: 0 = not contracted, 1 = never, 3 = occasional, 5 = daily use. Among those who already use AI.

Among respondents with at least one valid AI tool answer, 38/53 firms (71.7%) and 14/25 departments (56.0%) report daily use of at least one tool. This does not mean daily use of every tool. Daily research use in firms is 40.4% (21/52), compared with 3.8% for predictive case analysis (2/52).

What impact those who use it perceive

Perceived impact of AI (law firms vs. legal departments)
Law firmsLegal departments
AI lets us offer new legal services
3.21
3.18
AI lets us offer more accurate services
3.75
3.82
The share of time spent on routine tasks has decreased thanks to AI
3.92
3.61
AI tools have improved our capacity to handle more cases
3.60
3.54
At my firm we understand well enough how to implement AI
3.40
3.29

The dotted line marks the neutral point (3 = neither agree nor disagree).

What relates to perceiving greater impact from AI (Spearman's ρ)

Law firms

Adoption Index
0.64
Client experience
0.60
Efficiency
0.54
Team knowledge
0.53
AI usage
0.44
Tool usage
0.41
KPI maturity
0.38
Training
0.31
Infrastructure
0.30
Digitalization (%)
0.26
Ethical practices
0.26
Professional adaptation
0.13

Legal departments

AI usage
0.69
Adoption Index
0.62
Client experience
0.56
Training
0.48
Ethical practices
0.48
Efficiency
0.43
Tool usage
0.38
Team knowledge
0.35
KPI maturity
0.20
Infrastructure
0.14
Professional adaptation
0.04
Digitalization (%)
-0.04

In legal departments, one factor stands out above all: how much AI tools are used in daily work. Departments that use them daily report far greater impact than those that have them contracted but barely open them. In law firms there is no dominant factor. Client satisfaction, efficiency and the team's knowledge weigh more or less the same.

Perceived AI impact is associated with more frequent ethical practices in departments (ρ=0.477). The estimated association in firms is smaller (ρ=0.256) and not statistically significant. This does not establish absence of a relationship or absence of controls.

What distinguishes those who have already adopted AI?

The firms that have already implemented it stand out in two things: they feel more efficient and their team knows more about technology. The differences are clear, not marginal. Adopting AI could be the result of a prior path: first the team learns, the firm organizes its operations, and the decision emerges from there.

In legal departments we found nothing that distinguishes those that adopted AI from those that did not. Neither size, team knowledge nor perceived efficiency. Adoption is evenly distributed. This indicates that in a corporate legal department the decision could come from outside, a company directive or a corporate licence that has already arrived.

Ethical framework

By ethical practices in the use of AI we mean four concrete and measurable behaviors: assessing ethical risks before implementing a tool, applying confidentiality protocols when using it, consulting professional codes of conduct in the face of dilemmas, and documenting and reviewing the ethical impact of its use. The scale measures how frequently they are carried out (from "never" to "always").

Frequency of ethical practices in the use of AI
Law firmsLegal departments
We assess ethical risks before implementing AI
3.87
2.86
We apply specific confidentiality protocols when using AI
3.83
3.50
We consult traditional codes of conduct when using AI
3.23
3.11
We document and review the ethical impact of using AI
3.13
2.46

The dotted line marks the neutral point (3 = half the time).

This chart should be read carefully so as not to reverse its meaning: it is the law firms that assess ethical risks most frequently (clearly above legal departments), not the other way around. In legal departments, these practices barely reach half the time.

What relates to more frequent ethical practices (Spearman's ρ)

Law firms

Adoption Index
0.54
KPI maturity
0.54
Infrastructure
0.44
Client experience
0.41
Team knowledge
0.41
Training
0.38
Tool usage
0.34
Efficiency
0.33
AI usage
0.33
Perceived impact
0.26
Professional adaptation
0.17
Digitalization (%)
0.16

Legal departments

Adoption Index
0.70
Training
0.63
Perceived impact
0.48
AI usage
0.47
Team knowledge
0.44
Client experience
0.42
Tool usage
0.39
KPI maturity
0.38
Infrastructure
0.28
Efficiency
0.26
Digitalization (%)
0.08
Professional adaptation
-0.28

In law firms, ethics seems to come from management maturity: what most accompanies having formal ethical protocols is measurement discipline. Technological infrastructure and the client relationship weigh more than training; training appears, but in fifth place. These are correlations, not causal relationships, but they inform us about the pattern.

In legal departments, the channel is training: here it is indeed the dominant factor, with almost twice the strength seen in law firms. Ethical practices could also encourage training. What is clear is that this is consistent with how corporate functions work: compliance culture establishes practices through training programs.

The gap that matters: benefit without governance

The delicate finding is not one of levels, but of disconnection. In law firms, perceiving benefits from AI does not go hand in hand with having formalized ethical practices: one can be convinced of the value of AI without having established any controls over its use.

Before implementing AI, 34.0% of firms (18 of 53) and 53.6% of departments (15 of 28) assess its ethical risks at most half the time. These percentages refer to respondents to this question. This measures how often they review risks, not whether they have written AI policies.

In a context where the regulation of legal AI is still developing, this gap between perceived benefits and governance is a risk that deserves attention, both for law firms and legal departments.

What was tested and what did we find?

  • Do organizations assess ethical risks before using AI? In law firms, yes, with notable frequency. In legal departments, barely halfway. Ethical concern is unevenly institutionalized.
  • Are codes of conduct applied consistently in the face of technological dilemmas? There is no shared standard: very rigorous organizations coexist with others that almost never do so.

The statement of these hypotheses and their statistical detail (including the correction of their direction) are in the methodology appendix.

In summary

Most participating organizations already use AI. Among respondents to the tools block, 71.7% of firms and 56.0% of departments use at least one tool daily, although some tasks rely on AI much more than others.

The point of attention is governance: firms assess ethical risks before implementing AI more frequently than departments.

Finding AI useful, by itself, does not tell us whether its risks are routinely reviewed, but in both groups enthusiasm for AI runs ahead of the practices that should accompany it.

How to cite

Prince Tritto, P., Villa Berger, P., Salmon, L., Muñoz Ponce, M., & Limón González, J. B. (2026). Technology Maturity Index for the Legal Industry: pilot study (data from Mexico, 2025). Law and Artificial Intelligence Laboratory, Universidad Panamericana; KermaPartners; Lemontech. https://labderia.up.edu.mx/en/legaltech-report