upload_tii5m7qmi4p2fyhgvxtc3qahx5.pdf

21

DIGITAL DIMENSION OF MIGRANT SMUGGLING AND

ARTIFICIAL INTELLIGENCE: RISKS AND EUROPEAN

RESPONSES

Sebastian GEORGESCU, Ph. D candidate



Abstract:

The digital transformation of transnational organized crime has

significantly

changed

the

way

networks

involved

in

migrant

smuggling

operate.

Social media platforms, instant messaging applications and other digital tools are

already integrated into various stages of organizing and facilitating illegal border

crossings. The rapid development and growing accessibility of generative artificial

intelligence, however, introduce a new dimension to the phenomenon.

The

article

analyzes

the

transition

from

the

use

of

digital

tools

to

the

emerging

integration

of

artificial

intelligence,

distinguishing

between

already

documented

practices,

developing

trends

and

prospective

risks.

To

this

end,

an

analytical

model

based

on

five

functions

(5F-AI)

is

proposed:

expansion,

identification, interaction, concealment and amplification, through which artificial

intelligence

can

amplify

the

efficiency

and

scale

of

the

activities

carried

out

by

organized crime groups. In parallel, the article examines how AI tools can support

law

enforcement

authorities

and

border

management

structures

in

processing

large

volumes

of

information,

conducting

risk

analysis,

identifying

relevant

connections and developing investigative leads.

Particular attention is given to the relationship between migrant smuggling

and

human

trafficking,

as

well

as

to

the

legal

and

ethical

limits

on

the

use

of

artificial intelligence within the European regulatory framework.

The

study

argues

for

the

need

for

an

intelligence-led

response

model,

supported

by

technology

and

kept

under

human

control,

in

which

artificial

intelligence

complements,

without

replacing,

professional

expertise

and

human

decision-making.

Keywords:

migrant

smuggling;

illegal

migration;

AI;

organized

crime;

border management; OSINT; risk analysis.

DOI

10

.

56082/annalsarscimilit.2026.3

.21

Introduction

Migrant smuggling, as a form of transnational crime closely linked

to the dynamics of illegal migration, is one of the fields in which adaptation

to

political,

technological

and

operational

change

is

particularly

rapid.

1

Networks

involved

in

facilitating

illegal

border

crossings

do

not

operate

exclusively

through

classic

hierarchical

structures;

Paolo

Campana



CAROL I National Defence University, email: sebastian.georgescu13@yahoo.com.

1

Europol,

Criminal

Networks

in

Migrant

Smuggling,

Europol

Spotlight

Report,

Publications

Office

of

the

European

Union,

Luxembourg,

2023,

pp.

2-3,

DOI:

10.2813/550319.

DIGITAL DIMENSION OF MIGRANT SMUGGLING AND ARTIFICIAL

INTELLIGENCE: RISKS AND EUROPEAN RESPONSES

22

highlights

the

fragmented

nature

of

the

activities

associated

with

migrant

smuggling, the relative ease with which new actors can become involved,

and

the

frequent

presence

of

small-scale

structures

with

local

reach

and

limited

hierarchical

development.

2

European

investigations

describe

a

flexible

ecosystem

made

up

of

organizers,

intermediaries,

recruiters,

transporters,

providers

of

counterfeit

documents

and

other

persons

who

provide logistical, financial or operational support. Europol underlines the

adaptable

nature

of

these

networks

and

their

capacity

to

make

use

of

legitimate firms and other lawful commercial activities to support logistics

and conceal illegal activities.

3

In this context, the digital dimension can no longer be regarded as a

secondary element of how groups involved in migrant smuggling operate.

Social media platforms are used to promote services, attract migrants and

recruit facilitators, while instant messaging applications are frequently used

at

a later stage, for the

exchange of operational information

and

to

move

communications

away

from

public

channels.

4

The

role

of

digital

technologies has been highlighted by the relevant literature for more than a

decade,

with

research

showing

that

social

media

platforms

can

alter

the

structure

of

migration

networks

and

access

to

the

information

needed

for

movement.

5

Recent

research

on

illegal

crossings

of

European

borders

confirms the existence of genuine digital infrastructures in which migrants,

facilitators and other actors exchange information on routes and modes of

travel.

6

In March 2026, on the occasion of the launch of the new European

Centre against Migrant Smuggling (ECAMS), Europol underlined that the

activities

of

migrant

smuggling

networks

are

increasingly

taking

place

online, confirming the integration of the digital dimension into their modus

operandi.

7

The development of generative artificial intelligence may represent

the

of

this

transformation:

falling

costs,

reduced

technical

complexity and the capacity to generate content in multiple languages give

2

Paolo Campana, „Human Smuggling: Structure and Mechanisms”, Crime and Justice, vol.

49, 2020, pp. 471-519, DOI: 10.1086/708663.

3

Europol,

Tackling

Threats,

Addressing

Challenges

-

Europol's

Response

to

Migrant

Smuggling and Trafficking in Human Beings in 2023 and Onwards, Luxembourg, 2024,

DOI: 10.2813/203212, pp. 3-4, 7-10, 12.

4

Europol, Tackling Threats...,

op. cit

., p. 12.

5

Rianne Dekker, Godfried Engbersen, „How Social Media Transform Migrant Networks

and

Facilitate

Migration”,

Global

Networks,

vol.

14,

no.

4,

2014,

pp.

401-418,

DOI:

10.1111/glob.12040.

6

Ismail Oubad, Rassa Ghaffari, „The Digital Infrastructures of Illegal Border Crossings:

Solidarity

Actors

and

Networks

in

the

Arabic

and

Persian

Speaking

Virtual

Spheres”,

Critical Criminology, vol. 33, 2025, pp. 71-88, DOI: 10.1007/s10612-025-09824-5.

7

Europol, „The Grip on Migrant Smuggling Tightens: New Department at Europol to Step

Up the Fight”, March 24, 2026.

Sebastian GEORGESCU, Ph.D candidate

23

networks tools that can increase the efficiency of their operations.

8

Europol

shows

that

AI

and

other

emerging

technologies

accelerate

criminal

activities, increasing their speed and reach.

9

However, a nuanced approach

is

required:

while

the

use

of

digital

platforms

by

migrant

smuggling

networks

is

already

well

documented,

the

use

of

AI

is

only

beginning

to

emerge as a distinct trend. UNODC flags, among the emerging challenges,

facilitators'

use

of

new

information

and

communication

technologies,

including documents generated with the help of artificial intelligence.

10

In

recent

academic

literature,

the

integration

of

AI

into

criminal

phenomena associated with illegal migration is analyzed explicitly mainly in

relation to human trafficking, where the ways in which AI can facilitate the

identification and contacting of potential victims, influence them during the

recruitment

process,

and

amplify

the

operational

capacity

of

criminal

networks

are

already

being

examined.

11

This

research

offers

a

relevant

comparative

benchmark,

without

its

findings

being

automatically

transferable

to

migrant

smuggling.

Therefore,

the

question

is

no

longer

whether the digital environment is used in migrant smuggling, this reality is

already

well

documented,

but

rather

the

extent

to

which

the

digital

infrastructure already integrated into the activity of these groups creates the

conditions for the use of AI at various stages of the facilitation activity and,

conversely,

how

authorities

can

use

these

technologies

to

identify,

investigate and dismantle the networks.

The study's hypothesis is that the use of artificial intelligence does

not lead to the emergence of a distinct phenomenon within illegal migration,

nor to a new manifestation of migrant smuggling, but rather amplifies the

efficiency,

speed

and

scale

of

practices

already

used

by

organized

crime

groups involved in organizing and facilitating illegal border crossings. This

may

extend

these

groups'

capacity

to

reach

persons

interested

in

illegal

border crossing, may accelerate communication, reduce language barriers,

automate

certain

activities

and

increase

the

level

of

sophistication

of

fraudulent

or

deceptive

content,

while

also

facilitating

the

conduct

of

activities across multiple geographic and linguistic spaces.

Conversely, the same tools can support law enforcement authorities

in

making

use

of

large

and

heterogeneous

volumes

of

information,

by

identifying patterns, connections and anomalies that are difficult to detect

8

OSCE/Bali Process, New Frontiers: The Use of Generative AI to Facilitate Trafficking in

Persons, Policy Brief, 2024, pp. 6, 11-12.

9

Europol, EU

-

SOCTA 2025, Luxembourg, 2025, p. 11.

10

UNODC, Legislative and Policy Solutions to Emerging Issues Related to the Smuggling

of Migrants, CTOC/COP/WG.7/2025/3, August 5, 2025.

11

Joel

Levesque,

„The

AI-Enhanced

Trafficking

Threat:

Examining

the

Technological

Evolution

of

Trafficking

in

Persons

Operations

with

AI

Tools”,

Journal

of

Human

Trafficking, 2025.

DIGITAL DIMENSION OF MIGRANT SMUGGLING AND ARTIFICIAL

INTELLIGENCE: RISKS AND EUROPEAN RESPONSES

24

through exclusively traditional methods. The use of AI can thus contribute

to guiding risk analysis, prioritizing operationally relevant information and

developing investigative leads, without the automatically generated results

replacing

professional

judgment

and

human

decision-making.

In

this

dual

use

of

artificial

intelligence,

as

a

multiplier

of

the

capacity

of

networks

involved in migrant smuggling and, at the same time, as a support tool for

authorities, one of the main future challenges for border management takes

shape.

1. Methodology and Conceptual Distinctions

The study uses a qualitative approach, based on document analysis

and

comparative

analysis,

supplemented

by

a

prospective

analysis

of

emerging

trends

and

threats.

12

The

main

documentary

basis

consists

of

assessments

and

documents

produced

by

Europol,

UNODC,

Frontex,

OSCE, INTERPOL, the European Commission and other EU institutions,

supplemented

by

recent

academic

literature

on

migrant

smuggling,

the

digital dimension of migration and the use of new technologies in border

management. Recent research on the integration of AI into related criminal

activities,

particularly

human

trafficking,

is

also

used

as

a

comparative

benchmark

for

identifying

possible

developments

in

migrant

smuggling,

without

automatically

transferring

conclusions

between

legally

distinct

phenomena.

To

limit

unwarranted

extrapolation,

information

on

the

use

of

technology by organized crime groups is analyzed at three levels. The first

level covers documented practices, supported by operational information or

empirical research, such as the use of social media platforms, instant

and

encrypted

messaging

applications,

geolocation

tools

and

counterfeit

documents. The second level concerns emerging uses of AI, already flagged

by

authorities

and

international

organizations

but

still

insufficiently

documented through public data to be considered widespread practices. The

third

level

concerns

prospective

risks,

namely

technically

plausible

uses

resulting from combining AI's current capabilities with already documented

modes

of

operation,

without

these

being

presented

as

already

confirmed

practices.

This

distinction

is

important

because

artificial

intelligence

should

not be confused with digitalization in the broad sense. Encrypted messaging,

cryptocurrencies,

social

media

platforms

and

navigation

applications

are

digital tools, but they do not, in themselves, constitute AI systems.

13

They

form

part

of

the

digital

infrastructure

already

used

in

migrant

smuggling

activities,

into

which

AI-based

functions

or

tools

may

subsequently

be

12

Glenn A. Bowen, „Document Analysis as a Qualitative Research Method”, Qualitative

Research Journal, vol. 9, no. 2, 2009, pp. 27-40.

13

Regulation (EU) 2024/1689 (Artificial Intelligence Act), Art. 3(1).

Sebastian GEORGESCU, Ph.D candidate

25

integrated.

The

distinction

between

the

broader

integration

of

digital

technologies and the use of AI to amplify or automate certain components

of these activities is therefore purely analytical in nature and does not imply

the existence of legally distinct forms of the offence.

The

assessment

of

emerging

trends

and

threats

also

takes

into

account

the

risk-analysis

principles

used

in

the

field

of

European

border

management,

namely

the

three-component

structure

of

the

Common

Integrated

Risk

Analysis

Model

(CIRAM),

without

the

study

seeking

to

apply

this

operational

methodology

in

full,

but

only

to

adopt

its

logic

of

graduated assessment of the threat, vulnerability and impact associated with

a security phenomenon.

14

2.

Migrant

Smuggling

and

Human

Trafficking:

Legal

Delimitation and Zone of Convergence

The Protocol against the Smuggling of Migrants by Land, Sea and

Air,

supplementing

the

United

Nations

Convention

against

Transnational

Organized

Crime,

constitutes

the

main

international

legal

instrument

on

migrant smuggling. Under Article 3(a) of the Protocol, migrant smuggling

involves facilitating the illegal entry of a person into a state party of which

that

person

is

not

a

national

or

permanent

resident,

for

the

purpose

of

obtaining,

directly

or

indirectly,

a

financial

or

other

material

benefit.

15

Human

trafficking

has

a

different

legal

structure.

Under

the

Protocol

to

Prevent,

Suppress

and

Punish

Trafficking

in

Persons,

Especially

Women

and

Children,

it

involves

the

recruitment,

transportation,

transfer,

harbouring or receipt of a person, through means such as threat, coercion,

fraud, deception or abuse of a position of vulnerability, for the purpose of

exploitation.

16

This distinction is essential for the present analysis, since the mere

fact

that

a

person

turns

to

a

group

involved

in

facilitating

illegal

border

crossing

does

not

confer

on

that

person

the

status

of

a

victim

of

human

trafficking.

In

migrant

smuggling,

the

initial

relationship

between

the

migrant

and the facilitator may be consensual

in

nature, with

the migrant

seeking illegal entry into another state and the facilitator seeking a financial

benefit. In human trafficking, the determining element is exploitation, and

the victim's consent

is

irrelevant

where the means set

out

in

the Protocol

have

been

used.

This

legal

delimitation

does

not,

however,

mean

that

a

person smuggled illegally is free from vulnerability. Europol highlights the

existence

of

connections

between

the

two

phenomena,

showing

that,

14

Frontex, Common Integrated Risk Analysis Model 2.1 (CIRAM 2.1), Warsaw, 2021.

15

Protocol against the Smuggling of Migrants by Land, Sea and Air, Art. 3(a), New York,

15 November 2000, ratified by Romania through Law No. 565/2002.

16

Protocol to Prevent, Suppress and Punish Trafficking in Persons, Art. 3(a), New York,

November 15, 2000.

DIGITAL DIMENSION OF MIGRANT SMUGGLING AND ARTIFICIAL

INTELLIGENCE: RISKS AND EUROPEAN RESPONSES

26

particularly in the case of long and costly journeys, migrants may end up

being exploited to pay off debts accumulated in connection with facilitation

services.

17

Specialized

literature

confirms

that

the

relatively

clear

legal

distinction between the two offences becomes much more difficult to apply

in

certain

concrete

situations.

Roxane

de

Massol

de

Rebetz

describes

the

existence

of

a

“grey

zone”

at

the

intersection

of

migrant

smuggling

and

human

trafficking,

in

which

consent,

vulnerability,

debt,

abuse

and

exploitation may overlap over the course of the same journey, with persons

who

voluntarily

begin

their

journey

later

ending

up

in

situations

of

exploitation.

18

From a border-control perspective, this zone of convergence

requires

an assessment

of the individual context

of the journey, since the

apparently

voluntary

nature

of

the

movement

or

the

existence

of

valid

documents does not, in itself, exclude the presence of elements of coercion,

deception or exploitation.

More

recent

research

also

examines

the

convergence

between

the

two phenomena in regional contexts in which the same vulnerability factors

and

the

same

organized

crime

structures

can

connect

the

facilitation

of

illegal migration with the exploitation of persons.

19

These findings are also

relevant to the analysis of how AI could amplify the risk of a shift from the

consensual

relationship

specific

to

migrant

smuggling

toward

forms

of

exploitation

specific

to

human

trafficking,

since

the

same

tools

for

personalizing communication and building credibility can be used in both

contexts,

even

though

the

legal

classification

of

the

activity

remains

different.

3. From Existing Digital Infrastructure to the Integration of AI

The

traditional

model,

based

mainly

on

direct

contact

with

local

intermediaries and physical coordination of transport, no longer reflects the

full diversity of current modes of operation. The digital environment enables

remote

coordination,

the

distribution

of

tasks

among

persons

located

in

different

states,

and

the

conduct

of

certain

components

of

the

activity

without direct contact

20

. Europol shows that digital tools are used at various

stages

of

migrant

smuggling,

from

promoting

routes

and

prices

and

communicating

with

migrants,

to

providing

travel

instructions

and

producing

counterfeit

documents,

with

leaders

coordinating

activities

without being present along the route, intermediaries managing recruitment,

17

Europol, Tackling Threats..., op. cit., p. 6.

18

Roxane

de

Massol

de

Rebetz,

„How

Useful

Is

the

Concept

of

Transit

Migration...”,

European Journal on Criminal Policy and Research, vol. 27, 2021, pp. 48-49.

19

Concepción Anguita-Olmedo, „The Convergence of Trafficking and Migrant Smuggling

in West Africa: Migration Pressure Factors and Criminal Actors”, Social Sciences, vol. 14,

no. 8, 2025, art. 447.

20

Europol, Criminal Networks in Migrant Smuggling,

op. cit

., pp. 3, 5.

Sebastian GEORGESCU, Ph.D candidate

27

logistics

and

payments,

and

other

persons

providing

transport,

supplying

false

documents

or

offering

other

forms

of

specialized

support.

21

A

2025

qualitative study shows that facilitators actively use social media platforms

to communicate with migrants, while the use of encrypted communications

represents a growing difficulty for law enforcement authorities.

22

Networks

also

run

sophisticated,

professional

online

campaigns

to

recruit

drivers

to

carry

out

transport,

providing

information

on

working

conditions,

pay,

vehicle insurance and the support promised in the event of detection by the

authorities.

23

Earlier

research

has

also

shown

that

the

online

environment

facilitates

the

identification

of

and

contact

with

facilitators

and

allows

migrants

to

compare

information

on

costs,

routes,

destinations

and

reputations,

effectively

turning

the

decision

to

use

a

facilitator

into

an

informed

process,

supported

by

reviews

and

recommendations

similar

to

those found in legitimate e-commerce.

24

Contacting persons interested in the

facilitation

of

illegal

border

crossing

is

no

longer

confined

to

the

communities or geographic locations in which intermediaries operate, and

different

components

of

the

activity

can

be

coordinated

remotely

and

distributed

among

persons

with

distinct

roles,

thereby

extending

the

geographic range in which networks can effectively operate.

This change is also recognized at the European institutional level. In

2024, Europol and the European Commission supported the development of

a

European

network

of

experts

capable

of

making

more

effective

use

of

information

obtained

from

the

online

environment,

a

direction

given

concrete form through DigiNeX, the network of digital investigators aimed

at

monitoring

open

sources

and

detecting

and

analyzing

online

activities

associated with migrant

smuggling.

25

Within the first joint actions carried

out

by

DigiNeX

in

October

2025,

investigations

focused

on

identifying

suspects'

digital

fingerprints,

mapping

the

links

between

the

persons

involved, and locating points relevant to investigations; the results presented

by

Europol

in

December

2025

included

160

social

media

accounts

associated

with

migrant

smuggling

activities,

which

were

assessed

and

21

Europol, Criminal Networks in Migrant Smuggling,

op. cit

., pp. 2-6.

22

Özlem

Özdemir,

Blent

Baykal,

Elif

Başak

Sarıoğlu,

„The

Use

of

Social

Media

in

Irregular Migration and Migrant Smuggling: A Qualitative Study in Trkiye”, Journal of

Borderlands Studies, vol. 40, no. 6, 2025, pp. 1497-1519.

23

Europol, Tackling Threats...,

op. cit

., p. 12.

24

Parisa Diba, Georgios Papanicolaou, Georgios A. Antonopoulos, „The Digital Routes of

Human

Smuggling?

Evidence

from

the

UK”,

Crime

Prevention

and

Community

Safety,

vol. 21, 2019, pp. 159-175.

25

Europol, „New Network to Target Migrant Smugglers in the Digital Domain”, May 3,

2024.

DIGITAL DIMENSION OF MIGRANT SMUGGLING AND ARTIFICIAL

INTELLIGENCE: RISKS AND EUROPEAN RESPONSES

28

referred

to

the

platforms

concerned

for

appropriate

action

to

be

taken.

26

INTERPOL

identifies

a

similar

trend,

showing

that

social

networks

facilitate

contact

and

the

exchange

of

information

between

migrants

and

facilitators, while instant messaging applications and real-time geolocation

technologies simplify the planning and execution of journeys.

27

As

a

result,

combating

migrant

smuggling

can

no

longer

be

addressed

solely

in

relation

to

the

physical

border;

from

an

investigative

perspective,

part

of

the

relevant

activity

takes

place

before

the

migrant

actually

reaches the border, in

the digital environment

where services are

promoted,

negotiated

and

coordinated.

Overall,

these

developments

show

that the integration of artificial intelligence into migrant smuggling does not

start

from

an

environment

devoid

of

technological

infrastructure.

Social

media

platforms,

encrypted

communications,

navigation

tools,

digital

payment

services

and

information

flows

are

already

present

in

various

components

of

the

criminal

activity,

and

AI

can

be

integrated

into

this

already

existing

infrastructure,

contributing

to

the

amplification

or

automation of functions previously performed by digital means.

4. Artificial Intelligence in the Dynamics of Migrant Smuggling:

The 5F-AI Model

To systematize the ways in which artificial intelligence can alter the

activity of networks involved in migrant smuggling, the article proposes an

analytical model based on five functions (5F-AI): expansion, identification,

interaction,

concealment

and

amplification.

These

functions

do

not

necessarily represent sequential stages; rather, they may overlap and occur

at

different

points

in

the

criminal

activity,

depending

on

the

network's

technical resources and level of sophistication.

The model is analytical rather than exhaustive or purely descriptive:

its purpose is not to document individual cases of AI use, but to provide a

common conceptual framework applicable both to threat analysis and to the

construction of coherent institutional responses. Each of the five functions is

examined

below

with

reference

to

its

level

of

available

documentation,

confirmed

practice,

emerging

trend

or

prospective

risk,

so

as

to

avoid

presenting technical possibilities as established facts.

4.1. Expansion - Extending Reach in the Digital Environment

The first function concerns the extension of reach. Generative AI can

rapidly produce content in the form of text, images, audio and video, and the

development

of

multilingual

capabilities

allows

communication

to

be

26

Europol,

„Firm

Stand

against

Migrant

Smuggling:

Global

Alliance

Convened

in

Brussels”, December 12, 2025.

27

INTERPOL, Project CCISOM: New Technologies - Cyber Challenges in Smuggling of

Migrants and Human Trafficking.

Sebastian GEORGESCU, Ph.D candidate

29

adapted

to

different

linguistic

and

cultural

groups.

28

By

reducing

these

barriers, AI tools can significantly extend the capacity of groups involved in

migrant smuggling to reach persons and communities that were previously

harder to access. In the field of human trafficking, OSCE already warns that

AI

can

amplify

the

phenomenon's

reach

and

complexity,

which

calls

for

coordinated

responses

at

the

national,

regional

and

international

levels.

Automatically transferring these conclusions to migrant smuggling would,

however, be unwarranted. The possibilities offered by these technologies are

nonetheless also relevant in this field, and Europol explicitly shows that AI

applications can create new opportunities for attracting persons interested in

illegal border crossing, including through disinformation campaigns and the

use of deepfake technology, which can lead a larger number of people to

turn

to

facilitators

29

.

Consequently,

automated,

multilingual

and

personalized content generation can extend the reach of groups involved in

migrant smuggling and their capacity to reach broader categories of persons.

4.2. Identification - Identifying and Selecting Target Groups

The

second

function

concerns

the

identification

and

selection

of

categories of persons online, without necessarily involving the construction

of an individual profile for each person. In the field of human trafficking,

OSCE

describes,

as

an

emerging

risk,

the

possibility

of

automating

the

identification of potential victims through the use of AI agents configured to

browse social networks and initiate interactions, as well as the possibility of

scaling up such activities.

30

By functional analogy, in migrant smuggling,

analysis of the language used, participation in certain online communities,

interest

in

certain

routes,

and

other

publicly

available

information

could

enable groups to target

their messages toward categories of persons more

likely to be interested in the facilitation of illegal border crossing. Such a

scenario

must

currently

be

treated

as

an

emerging

risk,

not

as

a

systematically

documented

practice

in

migrant

smuggling.

The

technological

relevance

of

this

possibility

is,

however,

confirmed

by

developments observed in other criminal activities carried out in the digital

environment:

Europol

shows

that

the

use

of

generative

AI

enables

the

personalization

of

messages

and

the

targeting

of

a

very

large

number

of

people at high speed.

31

The difference from conventional digital promotion

therefore lies in the ability to select and prioritize communities, languages,

themes

and

categories

of

interest,

and

subsequently

to

adapt

messages

accordingly.

28

OSCE/Bali Process, New Frontiers...,

op. cit

., pp. 6, 11-12, 16-19.

29

Europol, Tackling Threats...,

op. cit

., p. 13.

30

OSCE/Bali Process,

op. cit.,

pp. 13, 16-17.

31

Europol, IOCTA 2026, Luxembourg, 2026, p. 17.

DIGITAL DIMENSION OF MIGRANT SMUGGLING AND ARTIFICIAL

INTELLIGENCE: RISKS AND EUROPEAN RESPONSES

30

4.3. Interaction - Building Credibility Online

The

relationship

between

migrant

and

facilitator

is

strongly

influenced by the level of trust between the two parties. In the absence of

formal

mechanisms

to

guarantee

that

agreements

will

be

honoured,

the

relationship is marked by uncertainty and an information imbalance, which

increases the importance of trust and reputation in choosing a facilitator.

32

Generative

AI

can

influence

precisely

this

dimension

of

the

relationship:

conversational

assistants,

machine

translation

adapted

to

cultural

context,

the

generation

of

images

and

video

materials,

and

voice

synthesis

and

cloning can contribute to building digital identities and seemingly credible

interactions,

including

between

persons

who

do

not

speak

the

same

language.

33

Deepfake technologies and synthetic identities can amplify the

potential

for

deception.

Europol

assessments

of

biometric

system

vulnerabilities

highlight

the

possibility

of

using

them

to

present

a

person

under

a

different

identity

or

to

avoid

identification,

including

in

remote

verification processes.

34

In the specific field of migrant smuggling, Europol

explicitly

identifies

the

possibility

that

AI

applications

may

be

used

to

attract

persons

interested

in

illegal

border

crossing,

showing

that

disinformation

and

deepfakes

can

help

mobilize

migrants

in

an

irregular

situation and increase demand for the services offered by facilitators.

35

This

assessment

does

not

demonstrate

the

existence

of

systematic

use

of

deepfakes

in

migrant

smuggling,

but

it

shifts

the

analysis

from

a

purely

abstract technical possibility to an operational risk explicitly identified by

Europol. Within the 5F-AI model, the Interaction function thus denotes the

use of AI to strengthen the credibility of a message, personalize interaction,

and lower the trust barriers between the facilitator and the targeted person,

regardless of whether the means used consist of machine translation, visual

content generation or voice synthesis.

4.4. Concealment - Hindering Detection of Digital Activity

The fourth function concerns the capacity to adapt digital content in

an

environment

in

which

authorities

and

online

platforms

identify

and

remove

accounts

or

materials

associated

with

the

promotion

of

migrant

smuggling

activities.

Generative

AI

enables

the

large-scale

production

of

synthetic content and the automation of activities that previously required

human intervention, which can facilitate the repeated alteration of wording,

32

Paolo Campana, Loraine Gelsthorpe, „Choosing a Smuggler: Decision-making Amongst

Migrants Smuggled to Europe”, European Journal on Criminal Policy and Research, vol.

27, 2021, pp. 9-13.

33

OSCE/Bali Process,

op. cit

., pp. 13, 16-19.

34

Europol, Biometric Vulnerabilities...,

op. cit

., pp. 39-41.

35

Europol, Tackling Threats...,

op. cit

., p. 13.

Sebastian GEORGESCU, Ph.D candidate

31

images or the style of

messages.

36

In the

case of other criminal

activities

carried out in the digital environment, the accessibility of these technologies

is already associated with additional difficulties in identifying content and

the persons generating it. In the case of migrant smuggling, the use of AI for

this

purpose

must

currently

be

treated

as

an

emerging

risk,

not

as

a

systematically

documented

practice.

From

an

operational

standpoint,

the

ability

to

rapidly

generate

variants

of

the

same

content

can

reduce

the

effectiveness

of

mechanisms

based

solely

on

the

recognition

of

known

terms, images or patterns. The relevance of this challenge is also recognized

at the political level. The action plan adopted by the G7 interior and security

ministers in October 2024 calls for cooperation with social media platforms

to

monitor

and

remove

content

promoting

migrant-smuggling-related

services and explicitly recommends the use of new technologies, including

AI tools, for the rapid removal of online activities through which facilitators

promote illegal migration.

37

Within the 5F-AI model, Concealment does not

denote the achievement

of anonymity,

but

rather the potential capacity to

rapidly adapt digital presence and content in order to hinder identification

and sustain online activity.

4.5. Amplification - Scaling and Automating Activities

The final function concerns the possibility of expanding the volume

and pace of activities carried out with the support of AI. Its effect does not

necessarily stem from the emergence of entirely new criminal methods, but

from

the

capacity to

automate, reproduce and

rapidly adapt

practices that

already

exist,

reducing

the

time

and

resources

needed

to

carry

them

out.

Operationally, generative tools can enable the simultaneous management of

a

larger

number

of

interactions,

the

rapid

production

and

adaptation

of

content

in

multiple

languages,

the

synthesis

of

large

volumes

of

information,

and

the

automation

of

repetitive

communication

tasks.

38

Through

these

mechanisms,

activities

that

previously

required

more

extensive human involvement can be carried out with significantly greater

capacity and speed. In its 2026 analysis of cybercrime, Europol notes that

the

growing

accessibility

of

AI

tools

lowers

barriers

to

entry

and

allows

criminals to expand their activities more effectively, increasing their speed,

efficiency and reach.

39

Within the 5F-AI model, the Amplification function

thus

denotes

AI's

capacity

to

extend

and

accelerate

practices

that

already

exist, not to generate fundamentally new forms of criminality.

36

Europol, „Firm Stand against Migrant Smuggling”, December 12, 2025.

37

G7 Interior and Security Ministers, Action Plan to Prevent and Counter the Smuggling of

Migrants, October 2024.

38

OSCE/Bali Process,

op. cit

., pp. 11-12, 16-19.

39

Europol, IOCTA 2026, Luxembourg, 2026, p. 14.

DIGITAL DIMENSION OF MIGRANT SMUGGLING AND ARTIFICIAL

INTELLIGENCE: RISKS AND EUROPEAN RESPONSES

32

The five functions of the 5F-AI model should not be interpreted in

isolation:

expanding

reach

facilitates

the

identification

of

larger

target

groups; identification supports the personalization of interaction; interaction

strengthens

the

credibility

needed

to

sustain

the

activity;

concealment

protects

the

continuity

of

digital

presence;

and

amplification

links

all

of

these functions into a cycle that can sustain and expand itself over time, as

the network accumulates operational experience in using AI tools.

5. Document Fraud and Digital Identity

Document

fraud

has

long

been

one

of

the

tools

used

to

facilitate

illegal migration, and its digital dimension is becoming increasingly evident.

In 2026, Europol documented the existence of an online platform supplying

forged identity and administrative documents, in both physical and digital

formats,

used

to

facilitate

migrant

smuggling,

evade

border

checks

and

support

secondary

movements

within

the

European

Union.

40

The

development

of

generative

AI

adds

a

new

dimension

to

this

problem:

current

tools

allow

the

production

and

manipulation

of

highly

realistic

images,

including

the

generation

of

synthetic

faces,

the

alteration

of

a

person's features, or the creation of deepfakes, technologies that can be used

to

present

a

person

under

a

false

identity

and

to

compromise

biometric

verification processes.

41

This

development

calls

for

further

reflection

on

how

border

authorities

cooperate

with

structures

specialized

in

combating

document

fraud, given that generative technologies lower the technical skill threshold

needed

to

produce

high-quality

forged

documents.

The

implication

for

border management therefore goes beyond the simple verification of a travel

document's

authenticity.

As

digital

identities

and

materials

become

increasingly difficult to distinguish from authentic ones, verification of the

document presented must be cross-checked with other relevant elements of

the journey and with information available in authorized systems. Europol

recommends an integrated approach to the biometric identification process,

from the initial registration of data to identity verification, in order to reduce

exploitable

vulnerabilities.

42

This

development

calls

for

more

rigorous

checks, but does not justify automatically interpreting every discrepancy as

an indication of fraud.

40

Europol, „Fake Document Factory Dismantled in Spain: Around 800 IDs Seized”, 4 June

2026.

41

Europol,

Biometric

Vulnerabilities:

Ensuring

Future

Law

Enforcement

Preparedness,

Europol Innovation Lab Observatory Report, Luxembourg, 2025, DOI: 10.2813/8081090,

pp. 39-41.

42

Europol, Biometric Vulnerabilities...,

op. cit

., pp. 10-11, 55.

Sebastian GEORGESCU, Ph.D candidate

33

6.

Using

Artificial

Intelligence

to

Combat

Migrant

Smuggling:

From Information Overload to Stronger Analytical Capacity

The same technologies that can amplify migrant smuggling activities

can,

conversely,

provide

important

tools

for

law

enforcement

authorities.

One

of

the

current

problems

facing

investigations

is

not

a

lack

of

information,

but

the

volume,

diversity

and

speed

at

which

it

must

be

analyzed.

A

cross-border

investigation

may

involve

processing

a

considerable

volume

of

messages,

data

extracted

from

electronic

devices,

images, transactions, and information concerning persons, locations and the

relationships between them. Europol shows that AI can support the analysis

of such data volumes by classifying and filtering information,

identifying

patterns

and

connections,

extracting

relevant

elements,

and

automatically

translating communications into multiple languages.

43

The applicability of

these

technologies

to

border

management

is

no

longer

purely

theoretical.

EU-funded research projects have already tested the use of AI in maritime

surveillance:

COMPASS2020

used

AI-assisted

surveillance

systems

to

improve

operational

situational

awareness,

EFFECTOR

focused

on

supporting

decision-making

in

maritime

surveillance,

and

PROMENADE

developed automated solutions for vessel tracking and anomaly detection.

44

Risk

analysis

is

of

particular

importance.

Frontex

anticipates

the

strengthening

of

intelligence-led

activities

and

risk

profiling

in

border

management, alongside the expanding use of biometrics and AI.

45

The use

of such tools can facilitate the aggregation and prioritization of indicators,

but the distinction between identifying a risk and formulating a conclusion

about the person assessed must be maintained. Europol draws attention to

the risk of excessive reliance on the results produced by automated systems

and to the possibility of unwarranted suspicions arising when these tools are

used outside the purpose for which they were designed.

46

Such a process

does not transfer decision-making authority to automated systems, but rather

uses AI to manage information volume and support analytical work.

The expanding use of biometrics also raises questions regarding the

interoperability of existing European systems, such as EURODAC and the

Entry/Exit System (EES), with any AI-assisted analysis tools developed at

the national level, so that data correlation complies with both the technical

43

Europol, AI and Policing: The Benefits and Challenges of Artificial Intelligence for Law

Enforcement, Europol Innovation Lab, Luxembourg, 2024, DOI: 10.2813/0321023, pp. 11-

12, 16-21.

44

European

Commission

-

CORDIS,

Counteracting

Migrant

Smuggling:

A

Multifaceted

Approach to Fight Migrant Smuggling, Results Pack, 22 November 2024.

45

Frontex, Strategic Risk Analysis Report 2024, Warsaw, 2024, pp. 27, 47-48.

46

Europol,

AI

Bias

in

Law

Enforcement

-

A

Practical

Guide,

Europol

Innovation

Lab,

2025, pp. 18-19, 25.

DIGITAL DIMENSION OF MIGRANT SMUGGLING AND ARTIFICIAL

INTELLIGENCE: RISKS AND EUROPEAN RESPONSES

34

requirements

of

interoperability

and

the

legal

limits

on

the

purpose

for

which each dataset was originally collected.

This

technical

interoperability

cannot,

however,

replace

genuine

institutional

cooperation:

AI-assisted

analysis

tools

produce

operational

value only to the extent that the structures using them have a clear mandate,

common

working

procedures,

and

functional

channels

for

exchanging

results

with

the

other

institutions

involved

in

the

decision-making

chain,

from the initial detection of a risk to the concrete measure applied by the

competent authority.

These functions can be integrated into a possible operational cycle

for

the

use

of

AI

in

support

of

combating

migrant

smuggling,

structured

around five stages: (1) Detection - identifying relevant signals within a large

volume

of

information;

(2)

Correlation

-

establishing

links

between

apparently

disparate

pieces

of

information;

(3)

Prioritization

-

directing

resources toward information with high operational value; (4) Intervention -

the decision on and application of measures by the competent authorities, in

accordance

with

the

law;

(5)

Learning

and

Adaptation

-

incorporating

operational results to improve analysis and reduce errors. This cycle allows

each intervention to

feed subsequent

rounds of analysis, without the final

decision ever being automatically delegated to the system.

7.

Legal

Limits

on

the

Use

of

Artificial

Intelligence

in

Border

Management

Regulation

(EU)

2024/1689

on

artificial

intelligence

establishes

a

framework

directly

relevant

to

the

use

of

such

systems

in

the

field

of

migration,

asylum

and

border

management.

Certain

systems

intended

to

assess the risks posed by persons entering the territory of a member state, to

assist in the examination of applications for asylum, visas or residence, and

to detect, recognize or identify persons in this context are classified as high-

risk

systems.

47

The

Regulation

pays

particular

attention

to

the

accuracy,

non-discriminatory

nature

and

transparency

of

these

systems,

given

the

impact their use may have on the rights of the persons concerned. At the

same

time,

high-risk

systems

must

be

designed

so

that

they

can

be

effectively overseen by natural persons, and the results they produce must

be capable of being interpreted and, where necessary, disregarded, modified

or reversed. The Regulation also draws attention to the risk of automatic or

excessive

reliance

on

the

results

generated

by

these

systems.

48

From

a

border-management perspective, the implication is clear: the output of an AI

system

can

guide

analysis

or

trigger

additional

checks,

but

it

cannot

substitute for the legal conditions required for taking a decision or measure

concerning the person assessed. The use of the technology must therefore be

47

Regulation (EU) 2024/1689, Art. 6(2), Annex III(7) and Recital (60).

48

Regulation (EU) 2024/1689, Art. 14(1) and (4)(b)-(d).

Sebastian GEORGESCU, Ph.D candidate

35

assessed

both

in

terms

of

operational

efficiency

and

in

terms

of

the

possibility of verifying and justifying the result obtained.

The

Regulation

also

imposes

specific

governance

obligations

on

operators

of

high-risk

systems,

including

conducting

a

fundamental-rights

impact assessment before deployment, keeping activity logs that allow the

traceability

of

AI-assisted

decisions,

and

establishing

effective

human-

oversight

mechanisms

capable

of

identifying

and

correcting

the

system's

systemic errors.

For national authorities with responsibilities in border management,

these requirements are not merely additional administrative constraints, but

a

mandatory

framework

for

any

future

strategy

for

integrating

artificial

intelligence.

The

design

of

decision-support

tools

must

include,

from

the

design stage onward, mechanisms for human oversight, documentation and

verification of results, not merely the technical component of data analysis.

8.

Relevance

for

Romania

Following

Full

Integration

into

the

Schengen Area

The analysis of how AI can influence migrant smuggling cannot be

separated

from

the

specific

context

in

which

Romania

exercises

its

responsibilities as a member state with an external border of the Schengen

area, a context that heightens the stakes of the accuracy and timeliness of

the information available to national authorities.

Romania's full integration into the Schengen area also changes the

context in which illegal migration and migrant smuggling must be analyzed.

As of 1 January 2025, checks on persons at Romania's internal land borders

were

eliminated.

49

This

change

does

not

diminish

the

importance

of

combating

cross-border

crime;

rather,

it

increases

the

importance

of

risk

analysis, information exchange, police cooperation and activities carried out

at

the

external

border

and

within

the

territory.

Data

from

the

Romanian

Border

Police

for

2025

show

a

decrease

of

approximately

78%

in

the

number

of

persons

detected

during

illegal

crossings

compared

with

the

previous year, while 188 migrant smuggling offences were recorded during

the same period.

50

A decrease in detections cannot automatically be equated

with

the

disappearance

of

the

threat,

particularly

since

significant

recruitment, communication and coordination activities already take place in

the digital environment.

In this context, AI's relevance for Romania must

primarily be assessed in relation

to

the capacity to

make use of available

information: integrating analysis of the digital environment with operational

analysis,

the

investigation

of

migrant

smuggling

and

international

cooperation can enable the early identification of connections that cannot be

49

Council of the European Union, „Schengen: Council decides to lift land border controls

with Bulgaria and Romania”, December 12, 2024.

50

Romanian Border Police, „Annual Report of the Border Police for 2025”, March 3, 2026.

DIGITAL DIMENSION OF MIGRANT SMUGGLING AND ARTIFICIAL

INTELLIGENCE: RISKS AND EUROPEAN RESPONSES

36

revealed solely through physical border control. For Romania, the challenge

lies

not

only

in

adapting

the

control

apparatus,

but

also

in

strengthening

analytical

capacity,

so

that

available

information

can

be

effectively

correlated and used in a context in which cross-border crime is increasingly

connected to the digital environment.

Romania's

particular

position,

as

a

state

located

on

the

European

Union's external border and along active transit routes, combined with the

elimination of systematic checks at internal borders, requires a rebalancing

of

the

relationship

between

physical

control

and

intelligence

analysis,

through the progressive strengthening of the latter's role. Other states that

have recently undergone a similar process of eliminating systematic checks

at

internal

borders

offer

useful

benchmarks

for

anticipating

possible

developments:

their

experience

shows

that

the

absence

of

systematic

physical control does not reduce pressure from illegal migration, but only its

immediate

visibility,

shifting

the

centre

of

gravity

of

detection

from

the

border toward intelligence-based risk analysis and cross-border operational

cooperation.

In this context, strengthening national capacity to analyze the digital

environment, in conjunction with tools already developed at the European

level,

such

as

the

DigiNeX

network

or

projects

like

COMPASS2020,

EFFECTOR and PROMENADE, can serve as a concrete means by which

Romania can compensate, at the intelligence level, for the reduced physical

visibility of the illegal-migration phenomenon resulting from its Schengen

status.

This direction, in turn, requires clear inter-institutional coordination

among

the

structures

responsible

for

border

management,

combating

organized

crime

and

international

cooperation,

so

that

information

originating from the digital environment can be coherently integrated into

the

decision-making

process,

without

remaining

fragmented

among

institutions that, at present, largely act independently of one another.

CONCLUSIONS

The

increasingly

extensive

integration

of

digital

technologies

into

migrant smuggling activities is already well documented, whereas the use of

artificial

intelligence

in

this

context

is

still

emerging.

This

distinction

is

essential

for

interpreting

the

phenomenon:

technological

possibility

must

not

be

conflated

with

the

existence

of

widespread

and

consistently

documented use.

The

analysis

carried

out

supports

the

hypothesis

that

artificial

intelligence

does

not

generate

a

distinct

form

of

the

offence

of

migrant

smuggling, but can amplify and automate practices that already exist. The

proposed analytical model highlights precisely the functions in relation to

which

this

influence

may

become

relevant,

from

extending

reach

and

Sebastian GEORGESCU, Ph.D candidate

37

personalizing

interactions

to

adapting

digital

presence

and

increasing

the

volume of activities.

The

three-level

distinction,

documented

practices,

emerging

trends

and prospective risks, used throughout the study remains relevant for future

research in this field as well: as more operational data on the actual use of

AI

by

migrant

smuggling

networks

become

available,

some

of

the

risks

discussed here as emerging may be confirmed as widespread practices or,

conversely, shown not to have materialised at scale.

This

same

transformation,

however,

also

offers

authorities

the

opportunity to make more effective use of large and heterogeneous volumes

of

information.

The

value

of

AI

thus

derives

mainly

from

its

capacity

to

support analysis when the volume and complexity of information make it

difficult to quickly identify relevant elements. At the same time, the use of

these tools in border management cannot be separated from legal limits, the

protection

of

fundamental

rights,

and

the

need

to

verify

automatically

generated results. Technology can flag an anomaly, a connection or a risk,

but

its

significance

must

be

established

by

reference

to

context

and

the

information available.

For

Romania,

full

integration

into

the

Schengen

area

makes

this

development all the more relevant. The reduced role of systematic control at

internal borders, coupled with the persistence of risks associated with cross-

border

crime,

increases

the

importance

of

risk

analysis,

international

cooperation

and

the

use

of

information

from

the

digital

environment.

In

these

circumstances,

the

main

challenge

is

not

the

adoption

of

artificial

intelligence as an end in itself, but its integration into a border-management

model capable of harnessing the technological advantages without turning

algorithmic assessment into a substitute for professional analysis and legal

accountability.

A

relevant

direction

for

future

research

lies

in

the

empirical

validation

of

the

model

through

qualitative

methods

that

complement

documentary analysis, for example through interviews or focus groups with

practitioners

in

border

management,

the

fight

against

cross-border

crime,

particularly

migrant

smuggling, and European cooperation, who, through

their

direct

operational

experience,

can

validate,

refine

or

challenge

the

functions proposed by the 5F-AI model and their practical relevance to the

specific Romanian context.

Such empirical validation would also make it possible to transform

the

5F-AI

model

from

an

analytical

framework

into

a

foundation

for

building

a

national

inter-agency

strategy

in

response

to

the

digital

and

technological

dimension

of

migrant

smuggling,

adapted

to

Romania's

particular characteristics as a state on the external border of the Schengen

area.

In

our

view,

this

perspective

opens

up

a

natural

direction

for

continuing the present research.

DIGITAL DIMENSION OF MIGRANT SMUGGLING AND ARTIFICIAL

INTELLIGENCE: RISKS AND EUROPEAN RESPONSES

38

Ultimately, the effectiveness of using artificial intelligence in border

management

will

depend

on

institutions'

ability

to

turn

technological

advantage into genuine operational advantage, without diminishing human

oversight, legal accountability or the protection of fundamental rights.

BIBLIOGRAPHY:

1. Scholarly Works and Articles

ANGUITA-OLMEDO, Concepción, „The Convergence of Trafficking and

Migrant Smuggling in West Africa: Migration Pressure Factors

and Criminal Actors”, Social Sciences, vol. 14, no. 8, 2025, art.

447, DOI: 10.3390/socsci14080447.

BOWEN,

Glenn

A.,

„Document

Analysis

as

a

Qualitative

Research

Method”, Qualitative Research Journal, vol. 9, no. 2, 2009, pp.

27-40, DOI: 10.3316/QRJ0902027.

CAMPANA,

Paolo,

„Human

Smuggling:

Structure

and

Mechanisms”,

Crime

and

Justice,

vol.

49,

2020,

pp.

471-519,

DOI:

10.1086/708663.

CAMPANA,

Paolo;

GELSTHORPE,

Loraine,

„Choosing

a

Smuggler:

Decision-making

Amongst

Migrants

Smuggled

to

Europe”,

European

Journal

on

Criminal

Policy

and

Research,

vol.

27,

2021, DOI: 10.1007/s10610-020-09459-y.

DE

MASSOL

DE

REBETZ,

Roxane,

„How

Useful

Is

the

Concept

of

Transit

Migration

in

an

Intra-Schengen

Mobility

Context?

Diving

into

the

Migrant

Smuggling

and

Human

Trafficking

Nexus in

Search for Answers”, European Journal

on Criminal

Policy

and

Research,

vol.

27,

2021,

pp.

41-63,

DOI:

10.1007/s10610-020-09467-y.

DEKKER,

Rianne;

ENGBERSEN,

Godfried,

„How

Social

Media

Transform Migrant Networks and Facilitate Migration”, Global

Networks,

vol.

14,

no.

4,

2014,

pp.

401-418,

DOI:

10.1111/glob.12040.

DIBA, Parisa; PAPANICOLAOU, Georgios; ANTONOPOULOS, Georgios

A., „The Digital Routes of Human Smuggling? Evidence from

the

UK”,

Crime

Prevention

and

Community

Safety,

vol.

21,

2019, pp. 159-175, DOI: 10.1057/s41300-019-00060-y.

LEVESQUE,

Joel,

„The

AI-Enhanced

Trafficking

Threat:

Examining

the

Technological

Evolution

of

Trafficking

in

Persons

Operations

with

AI

Tools”,

Journal

of

Human

Trafficking,

2025,

DOI:

10.1080/23322705.2025.2572911.

Sebastian GEORGESCU, Ph.D candidate

39

OUBAD, Ismail; GHAFFARI, Rassa, „The Digital Infrastructures of Illegal

Border Crossings: Solidarity Actors and Networks in the Arabic

and

Persian

Speaking

Virtual

Spheres”,

Critical

Criminology,

vol. 33, 2025, pp. 71-88, DOI: 10.1007/s10612-025-09824-5.

ÖZDEMIR, Özlem; BAYKAL, Blent; SARIOĞLU, Elif Başak, „The Use

of Social Media in Irregular Migration and Migrant Smuggling:

A

Qualitative

Study

in

Trkiye”,

Journal

of

Borderlands

Studies,

vol.

40,

no.

6,

2025,

pp.

1497-1519,

DOI:

10.1080/08865655.2025.2504880.

2.

Documents

and

Reports

of

International

Organizations

and

Institutions

European Border and Coast Guard Agency (Frontex), Common Integrated

Risk Analysis Model 2.1 (CIRAM 2.1), Warsaw, 2021.

European

Border

and

Coast

Guard

Agency

(Frontex),

Strategic

Risk

Analysis Report 2024, Warsaw, 2024.

Europol,

Criminal

Networks

in

Migrant

Smuggling,

Europol

Spotlight

Report, Luxembourg, 2023, DOI: 10.2813/550319.

Europol, Tackling Threats, Addressing Challenges - Europol's Response to

Migrant Smuggling and Trafficking in Human Beings in 2023

and Onwards, Luxembourg, 2024, DOI: 10.2813/203212.

Europol,

AI

and

Policing:

The

Benefits

and

Challenges

of

Artificial

Intelligence

for

Law

Enforcement,

Europol

Innovation

Lab,

Luxembourg, 2024, DOI: 10.2813/0321023.

Europol,

The

Changing

DNA

of

Serious

and

Organised

Crime

-

EU-

SOCTA 2025, Luxembourg, 2025.

Europol,

Biometric

Vulnerabilities:

Ensuring

Future

Law

Enforcement

Preparedness,

Europol

Innovation

Lab

Observatory

Report,

Luxembourg, 2025, DOI: 10.2813/8081090.

Europol,

AI

Bias

in

Law

Enforcement

-

A

Practical

Guide,

Europol

Innovation Lab, 2025.

Europol, IOCTA 2026 - The Evolving Threat Landscape: How Encryption,

Proxies and AI Are Expanding Cybercrime, Luxembourg, 2026.

G7 Interior and Security Ministers, Action Plan to Prevent and Counter the

Smuggling of Migrants, Mirabella Eclano, 2-4 October 2024.

INTERPOL,

Project

CCISOM:

New

Technologies

-

Cyber

Challenges

in

Smuggling of Migrants and Human Trafficking.

OSCE Office of the Special Representative and Co-ordinator for Combating

Trafficking

in

Human

Beings;

Regional

Support

Office

of

the

Bali

Process,

New

Frontiers:

The

Use

of

Generative

Artificial

Intelligence

to

Facilitate

Trafficking

in

Persons,

Policy

Brief,

2024, ISBN 978-3-903128-81-1.

United

Nations

Office

on

Drugs

and

Crime

(UNODC),

Trafficking

in

Persons

&

Smuggling

of

Migrants

-

Module

11:

DIGITAL DIMENSION OF MIGRANT SMUGGLING AND ARTIFICIAL

INTELLIGENCE: RISKS AND EUROPEAN RESPONSES

40

Differences

and

Commonalities,

Education

for

Justice

University Module Series.

United

Nations

Office

on

Drugs

and

Crime

(UNODC),

Legislative

and

Policy Solutions to Emerging Issues Related to the Smuggling

of

Migrants,

document

CTOC/COP/WG.7/2025/3,

5

August

2025.

3. Institutional Documents and Communications

Council of the European Union, „Schengen: Council Decides to Lift Land

Border

Controls

with

Bulgaria

and

Romania”,

12

December

2024.

European

Commission

-

CORDIS,

Counteracting

Migrant

Smuggling:

A

Multifaceted

Approach

to

Fight

Migrant

Smuggling,

Results

Pack, 22 November 2024.

Europol,

„New

Network

to

Target

Migrant

Smugglers

in

the

Digital

Domain”, 3 May 2024.

Europol,

„Firm

Stand

against

Migrant

Smuggling:

Global

Alliance

Convened in Brussels”, 12 December 2025.

Europol,

„The

Grip

on

Migrant

Smuggling

Tightens:

New

Department

at

Europol to Step Up the Fight”, 24 March 2026.

Europol,

„Fake

Document

Factory

Dismantled

in

Spain:

Around

800

IDs

Seized”, 4 June 2026.

Romanian Border Police, „Annual Report of the Romanian Border Police

for 2025”, 3 March 2026.

4. Legal Instruments and International Legal Acts

Protocol

against

the

Smuggling

of

Migrants

by

Land,

Sea

and

Air,

supplementing

the

United

Nations

Convention

against

Transnational Organized Crime, New York, 15 November 2000,

ratified by Romania through Law No. 565/2002.

Protocol to Prevent, Suppress and Punish Trafficking in Persons, Especially

Women and Children, New York, 15 November 2000, ratified

by Romania through Law No. 565/2002.

Regulation (EU) 2024/1689 of the European Parliament and of the Council

of

13

June

2024

laying

down

harmonised

rules

on

artificial

intelligence (Artificial Intelligence Act).

Directive (EU) 2024/1712 of the European Parliament and of the Council of

13 June 2024 amending Directive 2011/36/EU on preventing and

combating trafficking in human beings.