Proposal no. 3: development of a new legal aid system that is more efficient (that respects human rights…) and less expensive for countries, and of the three global Internet applications necessary to implement it worldwide. These applications would use, among other tools, AI to help (a) judges, who will adjudicate legal aid (LA) applications, to write their decisions, (b) lawyers assisting the poor to write their briefs, and (c) the poor to present their legal aid requests.
The project
consists of developing (1) a new, more efficient and less costly
legal aid (LA) system, based on the creation (a) of a group of civil
servant judges specializing in the adjudication of LA requests, and
(b) a group of civil servant lawyers specializing in legal aid
missions [these two groups would be under the hierarchical
responsibility of the State and the OHCHR]; (2) a classification and
codification of all types of cases tried each year around the world;
and (3) three global Internet applications necessary to implement
this new LA system in all countries wishing to use it.
The two
applications helping judges to adjudicate and manage LA requests,
and lawyers to defend and manage the cases of the poor, would also
make it possible to record (1) the time spent (a) adjudicating LA
requests (and possibly resolving cases amicably through mediation),
and (b) defending the cases of the poor, and (2) all costs
associated with adjudicating LA requests and defending the poor. The
third application would help the poor formulate their legal aid
requests.
The objectives of
the new system are, of course:
(1) to correct all the
imperfections of the French LA system;
(2) to avoid the
systematic destruction of the rights and freedoms of the poor;
(3)
to fight more effectively against corruption and the clogging of
justice;
(4) to make better use of the most advanced
technologies [AI…, see Task Force on Justice report,
PJ
no 83, ch. 5],
and to provide new functionalities allowing us (a) to optimize the
legal aid and justice systems, and (b) to fight more effectively
against organized and transnational crime and terrorism;
(5) to
transfer knowledge and advanced computer systems (and technologies)
to poor countries; and
(6) to allow rich countries to fulfill
part of their ODA obligations while solving one of their own
important problems, and, of course, to help all countries achieve the
SDGs (especially targets 1, 5, 8, 10, 16, and 17, PJ
no 3, no 5).
This solution has many advantages both at the level of the legal aid offices (LAO), which adjudicate legal aid requests, and at the level of the lawyers who defend the poor. I will summarize some of them here.
For the legal
aid office (LAO), having a single national LAO will make it possible
to have judges specialized (1) in adjudicating LA requests and (2)
in mediation, and judges (a) who help resolve cases through
mediation before they are presented to the courts or the
prosecutor’s office in criminal matters, and (b) who, if mediation
is not possible, monitor cases before the various courts and
throughout the proceedings.
This means that it will often be
the same judge who adjudicates the LA request for a first‑instance
procedure (TA,),
and then for a possible appeal (CAA,), a supreme court petition (CE,
CC, CCo), and even possibly a request to the ECHR. [The judge should
also try to resolve cases amicably before the appeal, the supreme
court petition, etc., whenever possible.]
Such an
organization should:
(a) help decrease the workload of the
various jurisdictions by using mediation as much as possible to
resolve disputes [22% of cases tried in France each year involve at
least one party receiving LA];
(b) allow the use of a unified
working methodology for judges (and advanced technologies);
(c)
simplify the adjudication of LA requests at higher levels of
jurisdiction, because the judges who will adjudicate LA requests for
an appeal or a supreme court petition will already have studied the
files and cases of the LA applicants when they adjudicated the LA
requests at lower levels;
(d) reduce or minimize the operating
cost of LAOs, in particular by pooling management and IT expenses
with other countries.
Another objective of the national LAO is to significantly improve the quality of LA decisions. LA judges can and should do what is currently planned but never done: conduct an instruction of the LA request (collect documents and information, hear the parties, etc.) to try to resolve cases amicably if possible, and, if not, render LA decisions that are well‑motivated and precise, reducing the risk of losing attorneys’ fees for the State and of having a poor defending himself alone in court.
Another important
advantage of creating a national LAO (and a group of lawyers
specializing in LA) is linked to:
(a) the improvement of our LA
and justice information system in general;
(b) the improvement
of the evaluation of LA costs [currently we cannot calculate the
total cost of the legal
aid system in France, nor can we determine the detailed costs
(management, fees, transport, etc.) or the cost of adjudicating legal
aid; in fact, we know almost
nothing, except that lawyers supposedly give billions of euros in
gifts to the State and the poor every year, which is of course
false];
(c) better coordination with other information systems
of the Ministry of Justice; and
(d) better use of
expenditure‑mitigation mechanisms.
The creation of a group of civil servant lawyers specializing in LA would first of all guarantee respect for the constitutional rights of the poor — which is currently not the case — while (1) minimizing total LA expenditure and management expenditure, and (2) maximizing (or optimizing) the use of expenditure‑mitigation mechanisms and revenues to reduce public expenditure.
Respect for the
constitutional rights of the poor will result from, among other
things:
(a) the establishment of a unified working methodology
for lawyers in charge of LA missions, a system to control the quality
of the work done by lawyers, and a computer system allowing the
recording and detailed monitoring of this work;
(b) the
possibility of having the work of lawyers with less than five years
of experience supervised by an experienced lawyer;
(c) the
allocation of more difficult cases to more experienced lawyers, and
therefore the possibility of having several levels of unit value,
which is impossible with the current (old) French LA system (and of
paying lawyers according to their skills and experience);
(d)
the creation of a grid of the time necessary to resolve the different
types of cases, much more precise than the one we have now (and which
takes into account the competence and experience of lawyers and the
factual and legal difficulties of cases);
(e) reducing conflicts
of interest linked to the use of independent lawyers; and
(f)
the fact that lawyers (paid regularly) will not have to advance money
to LA clients.
The optimization
of total LA expenditure will result from, among other things:
(a)
the possibility of precisely establishing the total cost of LA [all
management costs (including travel expenses, secretarial work, IT),
the costs of adjudicating LA requests (judges’ salaries, etc.), and
lawyers’ fees to the nearest cent], which the Court of Auditors
rightly recommends evaluating (R2
PJ 33 );
(b)
the possibility of pooling management expenses more effectively (IT,
etc.), in particular by developing the IT system to help lawyers
defend the poor, and the system for monitoring the work of lawyers
that I recommend, and by using advanced technologies (AI,) and
videoconferencing systems to communicate with courts and judges (as
is already happening in the US) and with offenders in prisons to
minimize costs and travel expenses;
(c) the possibility of
generating income with certain cases (by taking a percentage of the
compensation obtained, as is done in the US in certain cases) and
maximizing the use of expenditure‑mitigation mechanisms [more
mediation; more frequent reimbursements of LA by the losing
party];
(d) the possibility of simplifying the payment of
lawyers [in France, instead of 1 million LA assignments paid to more
than 25,000 lawyers, we would have 12 salaries per year paid to
around 8,000 lawyers].
The development of two global IT applications — (1) an application to manage LA requests and help judges adjudicate them, and (2) an application to help lawyers manage their LA cases — would allow us to record the time spent on each LA request and case by the LA judge of the national LAO and by the LA lawyer specializing in LA, and therefore calculate the average time that judges take to adjudicate an LA request and to resolve the poor’s cases through mediation, and that lawyers take to resolve the poor’s cases for each type of case (taking into account the skill and experience of lawyers and judges and the factual and legal difficulty of the cases).
The development of
these two global applications would also allow us to create:
(1)
an international classification and codification of all cases
presented to justice (in each country and each year);
(2)
databases (a) of cases tried and (b) of poor parties to cases,
including delinquents and criminals (which are essential to more
easily control the work done by LA judges and lawyers and to fight
against organized and transnational crime and terrorism);
(3)
management data (which are essential for improving our justice and
police systems, etc.).
It is therefore a cross‑pillar data
action that fits well into the UN Data Strategy (PJ
no 55).
If we want (1) to
verify and control the quality of the work done by LA judges and
lawyers, (2) to find the best way to optimize our justice and police
systems and better fight against corruption and the clogging of
justice, (3) to make useful comparisons with other countries, and (4)
to allow researchers and experts to analyze the harmful behavior of
delinquents (and criminals) and find the best ways to correct them,
we need to:
– save the documents related to each case,
–
record the time spent adjudicating cases, and
– have (a) a
classification and international codification of types of cases, (b)
a database of all LA cases, and (c) an accurate estimate of the
average time that LA judges spend adjudicating LA requests or
resolving each type of case through mediation, and that lawyers spend
defending each type of case (depending on the experience and
knowledge of the LA judge and lawyer, and the legal and factual
complexity of cases).
More generally,
and in the longer term, we need an accurate estimate of the average
time that all judges spend adjudicating claims for each type of case
(depending on the experience of the judge and the technical and
factual complexity of the cases).
And, of course, given the
considerable cost of justice, the possibility of reducing justice
costs by pooling significant management expenses (IT, etc.) and by
using AI is also an obvious advantage of developing the same computer
system used by a large number of countries.
There are many advantages to developing a global legal aid system. I have described several in my letters, and I will summarize a few of them here.
Some countries do not have the expertise or resources to develop such advanced AI systems, so there is an obvious advantage in developing a system (in a rich country with the necessary expertise) that could be used everywhere — including in poorer countries that do not yet have the required expertise.
Also, if we have just one system for all countries, it will be easier to control the behavior of this AI system (possible biases, etc.) than to control the behavior of 180 different systems developed separately in each country.
Several rich countries have already developed, or are currently developing, AI systems to help judges and prosecutors render their decisions, and to help lawyers. (For example: in China, “System 206” to analyze files and assist judges and prosecutors, and “Judge AI” to decide simple cases; in the US, Lex Machina and Ross Intelligence to analyze millions of legal decisions; in Estonia, “Judge AI” for small cases; in France, Case Law Analytics to predict chances of success based on previous cases; in the UK, AI tested to help draft judicial decisions; in Canada, AI tested to speed up judges’ work; in Russia, algorithms developed to analyze previous cases and suggest verdicts for similar cases.) We can therefore use their experience to build a better and more advanced system. Also, since several advanced countries are now beginning to develop their own AI systems, if we wait longer, billions of dollars or euros will be spent and wasted, and the solutions found may not be as efficient as they should be — and as they would be if we worked together to develop the best possible AI system.
The UN (OHCHR) is encouraging countries to respect human rights and is helping them do so. Developing the best possible AI system — to help judges and prosecutors render honest and well‑motivated decisions that respect human rights, and to help legal aid lawyers defend the poor — and then maintaining and updating this system over time, is one of the best ways for the UN to achieve this.