Constitutional-Legal Aspect of Creating Large Language Models: the Problem of Digital Inequality and Linguistic Discrimination
Journal of Digital Technologies and Law,
Journal Year:
2025,
Volume and Issue:
3(1), P. 89 - 107
Published: March 27, 2025
Objective
:
to
study
the
impact
of
digital
inequality
on
implementation
constitutional
human
rights;
identify
risks
linguistic
discrimination
associated
with
development
and
use
large
language
models.
Methods
formal-legal
comparative-legal
methods,
as
well
method
theoretical
modeling.
These
approaches
are
complemented
by
general
scientific
methods
cognition,
allowing
for
a
comprehensive
analysis
legal,
technological
social
aspects
issue.
Results
research
found
that,
in
relation
models,
arises
due
uneven
digitalization
languages
manifests
itself
limited
access
natural
processing
technology.
In
turn,
unequal
this
technology
can
negatively
affect
constitutionally
guaranteed
rights
be
viewed
from
viewpoint
equality
non-discrimination
concepts.
The
author
emphasizes
that
technologies
exacerbate
existing
economic
inequalities
create
new
forms
discrimination.
Scientific
novelty
hidden
indirect
analyzed
manifest
themselves
artificial
intelligence
systems,
especially
generative
While
direct
detected
predictive
algorithms,
models
more
subtle
but
no
less
significant
cumulative
effects.
effects
contribute
formation
stereotypes
areas
such
professional
activity,
gender
ethnicity.
also
draws
attention
fact
increasing
autonomy
intelligence,
traditional
detection
becoming
effective,
which
requires
regulation
methods.
Practical
significance
results
provide
basis
identifying
assessing
legal
products
using
processing.
This
contributes
improvement
field
technologies.
article
offers
recommendations
lawmakers,
regulators,
developers
aimed
at
minimizing
Language: Английский
Personal Data in Artificial Intelligence Systems: Natural Language Processing Technology
Journal of Digital Technologies and Law,
Journal Year:
2024,
Volume and Issue:
2(1), P. 123 - 140
Published: March 20, 2024
Objective:
to
conceptualize,
from
the
viewpoint
of
personal
data
protection
legislation,
development
natural
language
processing
technology,
identifying
possible
legal
barriers
such
and
directions
for
further
research
issue.
Methods:
is
based
on
general
scientific
methods
cognition,
along
with
which
formal-legal
comparative-legal
were
applied,
as
well
method
theoretical
modeling.
Results
:
it
was
found
that
observance
regime
in
technology
leads
a
conflict
between
private-legal
public-legal
interests,
which,
turn,
creates
obstacles
this
technology.
The
shortcomings
existing
order
are
shown,
namely,
insufficient
correspondence
technical
features
development.
This
may
lead
risks
excessive
regulation,
or,
contrary,
neglecting
critical
areas
require
protection.
Problems
qualifying
involved
outlined.
An
attempt
made
define
limits
ensuring
lawfulness
within
material,
temporal
territorial
effect
regulation
field
identified
legality.
author
touches
upon
possibility
using
consideration,
important
improvement
information
communication
industry.
Scientific
novelty
paper
supplements
discussion
by
artificial
intelligence
systems
an
analysis
latter
insufficiently
studied,
making
relevant
law,
relations
arising
around
systems,
assess
impact
Practical
relevance:
applied
aspects
problems
researched
results
obtained
can
be
used
improve
public
creation
intelligence,
identify
developers
digital
products
Language: Английский
