Innovations using Generative AI for media literacy and fact-checking in
the European Union
Innovaciones con IA generativa para alfabetización y vericación en la Unión
Europea
doxa.comunicación |
nº 41, pp. 489-509
| 489
July-December of 2025
ISSN: 1696-019X / e-ISSN: 2386-3978
How to cite this article:
Zafra Arroyo, A.; Sánchez González, M. and Sánchez Gonzales, H. M. (2025). Innovations using Generative AI for media literacy and fact-
checking in the European Union.
Doxa Comunicación
, 41, pp. 489-509.
https://doi.org/10.31921/doxacom.n41a2874
Ana Zafra Arroyo.
Journalism graduate from the University of Malaga (2021), awarded a Bachelor’s degree Extraordinary
Prize. She holds a Master’s degree in Journalism from
Akdeniz University (Turkey), and in EU Multilevel Systems of
Fundamental Rights from the UNED. She is currently a predoctoral researcher in the FPU program of the Spanish
Ministry of Science, Innovation and Universities, aliated with the University of Malaga’s Department of Journalism,
where she is working on her thesis on media literacy and European digital regulation. She has taught subjects such as
Photojournalism and has done research on projects focused on the impact of disinformation on journalistic content,
audiences, and routines. She has written for academic journals, book chapters, and international conferences. She has
participated in educational projects funded by FECYT and on OTRI contracts dealing with disinformation in the eld of
medicine. She completed a predoctoral research stay at the Faculty of Information Sciences at the Complutense University
in Madrid. Her areas of research include disinformation, the right to information, intercultural communication, and
European media policies..
University of Málaga, Spain
[email protected]
ORCID:
0009-0006-6881-0351
María Sánchez González
holds a degree (2003) and a PhD (2007) in Journalism from the University of Málaga. She
has been an associate professor in the Department of Journalism since early 2010, specialising in digital culture,
communication and online participation, visualisation, strategic planning, and innovation. She has been a researcher
on ocial projects since 2004, and has led/participated in several OTRI projects, as well as being the author of dozens of
papers in indexed journals and other open-license online publications and resources. She is working as an Innovation
Technician at the International University of Andalusia (since 2007), where she is head of teacher training programmes,
open online learning, and other projects on e-learning and digital skills. She also coordinates DataBeers Málaga, a
community dedicated to organising data talks. She has been a guest speaker on numerous data dissemination and
transfer initiatives, received several awards, led various educational innovation projects, taught postgraduate and
teacher training courses, and completed internships at a number of European and Latin American universities. She is
accredited as a Senior Lecturer (2020). A member of ECREA, IAMCR, AE-IC, and others.
Further information at
www.
cibermarikiya.com
University of Málaga, Spain
[email protected]
ORCID: 0000-0003-3053-0646
is content is published under Creative Commons Attribution Non-Commercial License. International License CC BY-NC 4.0
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
490 |
nº 41, pp. 489-509
|
July-December of 2025
Innovations using Generative AI for media literacy and fact-checking in the European Union
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Recibido: 24/12/2024 - Aceptado: 28/05/2025 - En edición: 04/06/2025 - Publicado: 01/07/2025
Resumen:
Este estudio analiza el uso de la inteligencia articial generativa (IA
-
Gen) y otras innovaciones emergentes en proyectos europeos nancia
-
dos por el European Media and Information Fund (EMIF), centrados en
la lucha contra la desinformación. Mediante un cuestionario enviado
a los responsables de estas iniciativas, se identicaron casos concretos
de integración de IAGen en tareas de vericación, automatización y
alfabetización mediática. Los resultados permiten detectar tenden
-
cias emergentes, como la aplicación de la IA a procesos de vericación
automatizada, gamicación educativa y desarrollo de herramientas
personalizadas. El trabajo es relevante en la medida en que supone un
ejercicio de rendición de cuentas, al analizar proyectos nanciados con
fondos públicos, y subraya la importancia de avanzar en políticas de
regulación, formación y uso ético de la IA. A pesar de que el número de
respuestas fue limitado –debido al carácter incipiente de la tecnología
en el periodo estudiado (2022-2023)–, el estudio ofrece una radiografía
pionera del papel de la IAGen en el ecosistema europeo de lucha contra
la desinformación.
Palabras clave:
Inteligencia articial generativa; desinformación; alfabetización
mediática; gamicación; vericación de información
Received: 24/12/2024 - Accepted: 28/05/2025 - Early access: 04/06/2025 - Published: 01/07/2025
Abstract:
is study analyses the use of generative articial intelligence
(GenAI) and other emerging innovations in European projects
funded by the European Media and Information Fund (EMIF),
focused on combatting disinformation. A questionnaire sent to the
heads of those initiatives identied specic cases of integration of
GenAI in fact-checking, automation, and media literacy tasks. e
results shine a light on emerging trends, such as the application of
AI to automated fact-checking processes, educational gamication,
and the development of tailor-made tools. e study is important
in that it represents an exercise in accountability by analysing
publicly-funded projects and underscores the importance of
driving forward policies for regulation, training, and the ethical
use of AI. Although the number of responses was limited
–due to the nascent nature of the technology during the period studied
(2022-2023)– the study provides a pioneering snapshot of GenAI’s role
in the European counter-disinformation ecosystem.
Keywords:
Generative articial intelligence; disinformation; media literacy;
gamication; fact-checking.
Hada M. Sánchez Gonzales.
PhD in Journalism, Director of the Communication Projects Laboratory (LabProCom)
and Vice-Dean of Communication and Institutional Relations at the Faculty of Communication of the University of
Seville. Accredited as a Senior Professor by ANECA, she has been granted a six-year research period by the National
Commission for the Evaluation of Research Activity (CNEAI). She is part of the assessment team of the National Agency
for Evaluation and Planning (ANEP). She is also a member of the Spanish Society of Journalism (SEP). She researches
digital journalism (bots/chatbots), transmedia narratives, professional proles, and user guides for social media, as well
as emotional connectivity with the audience.
University of Sevilla, Spain
[email protected]
ORCID:
0000-0001-5078-2184
1. Introduction
e swift emergence of articial intelligence, especially generative articial intelligence (GenAI), has sparked debates in areas
such as education, law, and ethics. A conceptual delimitation of the most relevant terms is called for to address the topic of
this paper. Generative articial intelligence, also known as generative AI or GenAI, is dened as a subset of machine learning
that, beyond merely analysing data, also produces human-like results in response to user input (Woolley, 2024). It thus has
the ability to create content while simultaneously imitating human emotional intelligence and creativity through Natural
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
doxa.comunicación |
nº 41, pp. 489-509
July-December of 2025
Ana Zafra Arroyo, María Sánchez González and Hada M. Sánchez Gonzales
ISSN: 1696-019X / e-ISSN: 2386-3978
| 491
Language Processing (NLP) so that computers can understand, manage, and generate human- like language (Bharti et al.,
2024).
Its introduction into other everyday activities has prompted both enthusiasm and distrust that bring into question human
ability to manage the risks associated with its use. Among those risks are deepfakes, which are synthetic creations of images,
videos, or audio using advanced articial intelligence (AI) techniques, particularly deep learning. Such contents are designed
to imitate real-world data, making them highly plausible and often indistinguishable from authentic media. e term
“deepfake” derives from “deep learning” and “fake,” and reects the technology’s ability to manipulate digital content for
various purposes, both benign and malicious (Herke, 2023) (Singh et al, 2023).
e techniques can be used maliciously. For example, a manipulated video of Nancy Pelosi, in which she appeared to be
intoxicated, went viral when Donald Trump shared it on Facebook, and the platform refused to remove it (Waterson, 2019).
e European Commission denes disinformation as “veriably false or misleading information that is created, presented,
and disseminated for economic gain or to intentionally deceive the public and which may cause public harm” (European
Commission, 2018).
Measures to curb information disorders in Europe have largely been focused on user education and protection through the
implementation of media literacy and fact-checking measures, among others (Garriga et al., 2024). Media literacy refers to
the ability to engage with media through active inquiry, critical analysis of media texts, and understanding of the factors that
shape them. It encompasses the skills of accessing, using, creating, analysing, and evaluating content (Scharrer et al., 2022).
Fact-checking refers to data verication, especially in journalism, and is the process of conrming the accuracy of statements
by retrieving relevant evidence from various sources (Miranda et al., 2019).
Recent advances in generative AI have enabled the development of tools capable of holding conversations with a high degree
of realism and uency, as well as generating content from instructions provided by a human. Platforms such as Chat GPT or
Google Play employ Natural Language Processing to compose a variety of written content, including articles, social media
posts, essays, code, and emails (Hetler, 2024).
However, this progress, coupled with inecient regulation of the major social platforms, poses serious challenges to the control
of disinformation. e use of EU laws to exercise such control raises some concerns as state intervention in content circulating
online has been linked to non- democratic states (Haggart et al., 2021). To this end, a decentralised or networked regulatory
model has been adopted (de Cock Buning, 2018) which implies a shift in the locus of state regulatory activity to other, multiple
venues (Black, 2008). From this perspective, private actors play a key role (Mungiu-Pipidi 2015; Sedelmeier 2012). Some
authors argue that self- regulatory mechanisms such as the Code of Practice on Disinformation are ineective when faced
with disinformation. ey point out that the transparency reports provided by the Code’s signatories have multiple limitations,
such as repetition, vague descriptions, and a lack of quality data, and that they are also often unavailable to researchers (Park
& Culloty, 2023).
Large technology companies such as X (formerly Twitter), Google or Facebook “due to their role as intermediary services
and lack of editorial responsibility, (...) are not generally subject to the obligations arising from media legislation” (Gamito,
2023: 2) in Europe. In this context, the willingness of some of these platforms to cooperate in the regulation of such aspects as
492 |
nº 41, pp. 489-509
|
July-December of 2025
Innovations using Generative AI for media literacy and fact-checking in the European Union
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
algorithmic transparency, echo chambers, or fake content seems to respond more to public relations strategies than to a real
commitment or a substantial change in attitude (Scott, 2018; Allyn, 2024).
Funding initiatives such as the European Media and Information Fund (EMIF) are emerging as a key alternative to support
researchers, fact-checkers, non-prots, and entities dedicated to investigating disinformation, strengthening media literacy,
and fact-checking. e fund, established by the European University Institute and the Calouste Gulbenkian Foundation, also
receives funding from multiple donors, including Google. Various calls for proposals with an independent selection process
aim to address problems posed by the new digital ecosystem related to the lack of freedom, transparency, pluralism, non-
discrimination, and diversity (EMIF, 2023).
e workload of these fact-checking platforms increased signicantly during the COVID-19 pandemic, which fostered the
development of technological tools for monitoring and verication (Sánchez et al, 2024) of political discourse in real time
(Adair, 2020; Gonzalo, 2023). e application of AI tools such as Natural Language Processing (NLP) to fact-checking procedures
has accelerated workows (Gutiérrez Caneda & Vázquez-Herrero, 2024). NLP is a subset of articial intelligence that allows
computers to understand written and spoken human language and results in more intuitive interactions between humans and
machines, improving communication and functionality in numerous applications (Machiraju & Modi, 2018).
According to the literature review that Montoro-Montarroso et al. conducted on articial intelligence and machine learning
(ML) techniques, a subeld of AI “that makes up a series of methods, techniques, and tools for building intelligent systems from
large volumes of data” (2023: 4), there are three principal ways in which they assist in disinformation verication processes.
(1) Automated classication, which trains machine learning models to distinguish between true and false information. (2)
Feature extraction, which analyses textual and contextual elements associated with disinformation. (3) Hybrid verication,
which combines technological tools with human intervention. However, the authors warn that such tools “are not yet widely
available and/or lack the maturity necessary for their employment by non- technical users” (2023:10).
Santos (2023) concludes that technologies such as NLP and ML require human participation for their use, and have various
ethical considerations, such as algorithmic opaqueness. Sánchez, Sánchez, and Alonso (2024) show that, despite the growing
use of AI—particularly in image verication—in European newsrooms and fact-checking environments, there is still a certain
decit in the establishment of policies that guide its integration with traditional journalistic practices. us, the use of these
technologies generally lacks regulations or operational frameworks that allow for ecient implementation or consider
journalistic ethics.
Despite the aforementioned challenges, initiatives based on these innovations have demonstrated their capacity to educate the
public about the creation, detection, and potential risks of disinformation. e European Union is striving to improve media
literacy by focusing on the implementation of curricular proposals and projects adapted to local cultural contexts (Tiernan et
al., 2023). Generative AI is being integrated into educational environments to increase students’ digital literacy and AI tools
such as chatbots can help improve research skills and promote critical reading and fact-checking (Wetzel & Kani, 2025).
Gamied learning environments that integrate immersive technologies have been shown to increase student engagement,
motivation, and self-ecacy, fostering more collaborative and meaningful learning (Lampropoulos & Kinshuk, 2024). e use
of existing innovative models—such as gamication or storytelling—in media literacy education increases their potential when
doxa.comunicación |
nº 41, pp. 489-509
July-December of 2025
Ana Zafra Arroyo, María Sánchez González and Hada M. Sánchez Gonzales
ISSN: 1696-019X / e-ISSN: 2386-3978
| 493
combined with disruptive technologies such as virtual reality, augmented reality, and GenAI. is synergy aids the creation
of interactive and multimedia content, something particularly appealing to younger audiences. Rzabayeva, Kassymova, and
Pratama (2024) maintain that gamication signicantly enhances student engagement by incorporating dynamics such as
scores, rewards, and dierent levels, turning learning into an interactive, motivating experience. Its combination with AI
allows content to be personalised, adapted in real time, and to provide immediate feedback, improving the understanding of
complex concepts and knowledge retention.
In this context, this paper addresses the role of GenAI and other emerging technologies in European projects funded by the
EMIF, an initiative that supports proposals focused on combating disinformation beyond fact-checking, including research,
media literacy, and technological development. Given the emerging nature of many of these projects and the limited resources
they often have, their development relies on technological innovation.
is paper sets out three specic objectives: rst, to identify how these initiatives use emerging technologies, including GenAI,
in their ght against disinformation; second, to analyse project leaders’ perceptions of the potential and risks specically
associated with GenAI; and nally, to explore any training or digital literacy initiatives they may be preparing concerning AI
and disinformation.
Once the objectives were dened, research questions for the study were then posed:
Q1. What emerging technologies, including generative articial intelligence, are being incorporated into the initiatives
analysed to combat disinformation, and for what specic purposes?
Q2. How do those responsible for these initiatives perceive the role of generative articial intelligence in combating
disinformation, and what benets and risks do they identify in its implementation?
Q3. What training or digital literacy activities on articial intelligence and disinformation are these initiatives carrying out,
and what audiences are they targeting?
2. Method
As mentioned above, this study focused on projects supported by the European Media and Information Fund (EMIF). e
analysis, carried out in several phases, made use of various research techniques. e rst phase identied projects approved
by the EMIF from the start of its activity in September 2022 to the most recent call available at the time of the analysis (June
2024), corresponding to its seventh edition.
e projects were found utilising public documentation of the resolutions for each call, ensuring the comprehensiveness of the
census. Both ongoing (six approved in the last call) and completed projects (included in the 2022/23 EMIF progress report) were
included. A total of 39 projects were identied (see table below), representing the entire corpus of initiatives supported by the
fund to date. Basic information was included for each project, in addition to contact information, to allow for contextualisation
and a comparison of results. Specically, project name, country, date of implementation, type of organisation, and the activity
it appears to be engaged in (this last point being compared with the information provided by the project leaders themselves
in the next phase).
494 |
nº 41, pp. 489-509
|
July-December of 2025
Innovations using Generative AI for media literacy and fact-checking in the European Union
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Table 1. Projects analysed
Project name
CountryDate
INDEX: Advancing Open Source Investigation for Fact- Checking in France and Europe
France
Jun. 2024 -
…..
A Fuoco. A multimedia project focused on climate change and disinformation narratives
Italy
Jun. 2024 -
….
e Open Source Munitions Portal
United
Kingdom
Jun. 2024-
…..
CALYPSO.GR
– Collaborative AnaLYsis, and exPosure of diSinfOrmation in Greece
Greece
Jun. 2024-
….
FEED | Fact-Checking food, diets and eating behaviour
Portugal
Jun. 2024 -
…..
From Slovenia to Croatia: Implementing a successful fact-checking model
Slovenia
Jun. 2024 -
.....
True info
Belgium
01/10/2022-
31/03/2023
Ukrainian news Agency to integrate fact-checked content with the help of AFP and APA
France
01/09/2022-
28/02/2023
Fact-Checking to Improve Media Coverage of the Arctic in Europe
United
Kingdom
01/11/2022-
30/04/2023
Fact-checking the Ukrainian War in the Black Sea Region
Romania
01/11/2022-
30/04/2023
Scale-up Science Fact-Checking in Spanish Podcasts and Youtube
Spain
01/10/2022-
30/06/2023
Tackling false narratives in anti-corruption and the rule of law media reporting
Slovakia
November 20, 2022 –
November 19, 2023
SciNet – Network Analysis Tool to automatically rate web domains on their scientic
credibility
France
01/10/2022-
30/09/2023
FactBoost: Strengthening the fact-checking skills of journalists
Romania
01/10/2022-
30/09/2023
<p>[Enlace de URL / hc (has AS)]</p>
doxa.comunicación |
nº 41, pp. 489-509
July-December of 2025
Ana Zafra Arroyo, María Sánchez González and Hada M. Sánchez Gonzales
ISSN: 1696-019X / e-ISSN: 2386-3978
| 495
Czech Elections in the Era of Disinformation: e 2023 Presidential Election
Czech
Republic
01/10/2022-
28/02/2023
Media Against Disinformation
Poland
01/05/2023–
31/10/2023
“FAST: Fact-checkers acting to stop early-detected disinformation campaigns”
Spain
01/23/2023-
12/31/2023
Lucid News & Investigations
Czech
Republic
01/05/2023-
29/02/2024
Exposing the Slovenian Disinformation Ecosystem
Slovenia
01/04/2023-
30/09/2023
Fake News War. Fighting the Fake News regarding the war in Ukraine
Romania
01/04/2023-
30/09/2023
“Desinformiert” season 1. An Information war
Germany
01/03/2023-
31/08/2023
Made for social media: Fighting disinformation with fact check graphics
Germany
01/05/2023-
29/02/2024
Made for social media: Fighting disinformation with fact check graphics
Spain
01/02/2023-
31/12/2023
OMEDIALITERACY” will present an overview of the challenges and opportunities of
media literacy policies in Europe
Spain
01/02/2023-
31/12/2023
Decoding the Disinformation Playbook
Germany
01/10/2022-
30/09/2023
“Vrb
ě
tice case study: Eective policies and international cooperation for countering
disinformation.”
Czech Republic
01/10/2022-
30/09/2023
Uisce Faoi alamh, ISD’s investigation into the online mis- and disinformation
ecosystem in Ireland
United
Kingdom
01/11/2022-
31/07/2023
StarsAustria
01/12/2022-
30/05/2024
Disinformation & Trust DISINFTRUST
Spain
01/01/2023-
30/06/2024
496 |
nº 41, pp. 489-509
|
July-December of 2025
Innovations using Generative AI for media literacy and fact-checking in the European Union
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Stars4Media – Factchecking: Strategy for Eastern European Media Resilience
Belgium
01/10/2022-
31/03/2024
LATIF Leveraging argument technology for impartial fact-checking
Italy
01/02/2023-
31/07/2024
Teacher education in digital and media literacy: providing guidance, resources and support
Belgium
01/01/2023-
30/06/2024
Media Literacy Leads: Digital wisdom in an age of online disinformation.
United
Kingdom
01/09/2022-
29/02/2024
269268 - Empowering critical thinking education in local communities
Lithuania
01/11/2022-
30/04/2024
e Empowering schools in self-regulation of Media and Information Literacy processes
Italy
01/01/2023-
30/06/2024
ESCAPE FAKE 2.0
Austria
01/09/2022-
29/02/2024
Yo-Media – Youngster’s Media Literacy in Times of
CrisisPortugal
01/12/2022-
31/05/2024
Using micro-learning to train educators - a cascade approach to media and
information literacy
Portugal
01/10/2022-
31/03/2024
FAKE KNOW MORE
Poland
01/01/2023-
30/06/2024
Source: created by the authors
e second phase of the research (July-September 2024) saw the design of an online questionnaire, project leaders were
contacted by email and invited to participate by lling it in. is analysis included, among other aspects, the following points:
–
Project characterisation. In addition to its title, name, contact information for the leaders, and the relevant EMIF call for pro
-
posals, the type of initiating organisation was considered rst (universities or research centres; NGOs or non-prot organisa
-
tions; companies or startups; media outlets, news agencies or fact-checkers; government agencies; or others). Second, parti
-
cipants were asked to self-classify their project from the following list (with the option of selecting several): information, news,
and content editing; fact-checking; literacy (training, resource development, etc.). ey were also asked to indicate, from
another list, the specic activity/activities they were involved in: use of Open Source Intelligence to do research using publicly
accessible resources; analysis of audiovisual material (podcasts, YouTube, TikTok, etc.); following social media, blog posts,
and other online posts; 3D digital reconstruction or creation of digital images for fact-checking; production of news, reports,
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
doxa.comunicación |
nº 41, pp. 489-509
July-December of 2025
Ana Zafra Arroyo, María Sánchez González and Hada M. Sánchez Gonzales
ISSN: 1696-019X / e-ISSN: 2386-3978
| 497
bulletins, or newsletters; production of videos or podcasts; creation of video games; drawing-up of technical/expert reports;
media literacy in schools, for citizens, or for other social sectors; training for news professionals; development of solutions
and tools based on AI, big data, or other disruptive technologies; software development/programming; or other. Finally, the
proles of the professionals who made up each of the groups were compiled (journalists; researchers; teachers; scientists with
expertise in health or other subject areas; programmers/computer developers; statisticians/analysts/data scientists or other
data-related proles; linguists; others).
–
General perceptions of articial intelligence in the ght against disinformation (possibilities/potential and current challen
-
ges), in the form of open-ended questions. As well as, in the case of project managers using AI for related tasks, perceptions of
a series of specic aspects, discussed below.
–
Use of articial intelligence in the project (yes/no). is is the last question for those who answered in the negative, while
for those who answered yes, several non-exclusive options or types of use are considered: a) use of AI tools in some or all of
the development phases of the projects or tasks related to their activity; b) use of AI tools for specic fact-checking tasks; c)
development of a proprietary AI tool or participation with other entities in its development; d) organisation/presentation of
training specic to GenAI or teaching how to use AI tools when carrying out literacy training to combat disinformation; e)
other cases (open-ended response).
–
Regarding the use of AI in any of the phases for tasks related to their projects, including fact- checking, those who reported
using it were also asked about the specic type(s) of tasks they had used AI for: applied research / data collection and analysis;
task automation (interview transcriptions, format conversions, etc.); content production (videos, photos, audios, graphics,
presentations, etc.); fact-checking / source comparisons; distribution of personalised content to audiences; user/audience
assistance (chatbots, virtual assistants, etc.); or other. When speaking specically of fact-checking processes and based on a
previous study published in Ámbitos (Sánchez, 2022) as well as others, they were also asked to indicate whether AI was used
for source checking; verication in the processing and interpretation of texts; checking of previously published phrases; chec
-
king of audio in text; image checking; and/or video checking. e questionnaire also included a question asking participants
to indicate their most common practices regarding the use of one or more tools for the same task, or their use of free tools
or paid versions. It asked participants whether they had any type of funding or agreement concerning such use. e ques
-
tionnaire asked participants whether the project had any type of AI-use policy or regulations. Participants were also asked to
indicate their level of agreement (Likert scale from 0-3) with specic questions (“AI makes it easier for fact-checkers and news
professionals to verify information”; “AI is a tool that allows citizens to check information for themselves”; “AI has an impact,
more than on the ght against disinformation, on the manipulation of news and the spread of fake news”; and “AI does not
yet have a real impact on journalism”).
–
Regarding AI-based projects/tools developed in-house and including both those created by the project team alone and tho
-
se created together with other actors/entities, project managers were given an opportunity to provide information on up
to three tools (requesting them to select the most relevant ones if they had developed more). Based on the questionnaire
developed for previous analyses of AI use by Ibero-American and European fact-checkers (Sánchez, Sánchez, & Martínez,
2022; Sánchez, Sánchez, & Alonso, 2024), basic data was collected on the tools: name; year of creation and current availability
498 |
nº 41, pp. 489-509
|
July-December of 2025
Innovations using Generative AI for media literacy and fact-checking in the European Union
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
(yes/no); openness/availability to audiences (dierentiating between open and free, only available to the media, and other
cases); and access URL. eir typology was also considered (microsite, browser extension, online application, bot, mobile
application, service, and other); and the tool’s orientation/usefulness with respect to journalistic/news-checking tasks, taking
the classication proposed by Prodigioso Volcán (2023) as a basis for the latter: automatic content generation; fact-checking
procedures; management and purging of massive databases; distribution of automated responses/content to audiences; and
other. Moreover, it was considered whether or not the tools were framed within strategic initiatives/projects, whether there
had been external collaboration in their development and, if so, involving what types of entities (foundations, public/private
universities, other educational institutions, various public bodies/administrations, other types of organisations/companies,
etc.). Questions were also asked about their funding; and their views of the external impact of each tool and its impact on the
project itself in terms of innovation was compiled.
–
e project leaders were also asked, regarding the training/literacy initiatives on AI and disinformation promoted by the
projects studied, based on our own previous work (Sánchez, Sánchez, & Martínez, 2022; Sánchez, Sánchez, & Alonso, 2024),
whether, on the one hand, aside from providing AI-focused training or not, they taught use of AI tools when teaching disin
-
formation literacy (in any of their activities or resources with a broader focus/content); and, on the other hand, whether the
project provided training specically focused on AI-related disinformation. Basic data was collected from the projects that
responded armatively to this second question concerning the most important training experiences (up to 3 per project).
First, the content of the experience, dierentiating between an introduction to AI tools, an assessment on the detection of di
-
sinformation with AI, fundamental ethical aspects of AI, checking and investigating false or manipulated information with AI,
or others. Second, the category (virtual; in-person; or hybrid). ird, type(s) of experience (courses/conferences; workshops;
seminars/webinars; tutorials/guides; reports/manuals/books; blog posts; content/social media activities; other). Fourth, for
-
mat (visual posters/infographics/images; videos; audios/podcasts; games; other). Fifth, target audience(s) (news professio
-
nals; professionals from other sectors; communication students; students from other sectors/educational levels; professors
of journalism/communication; professors from other sectors/educational levels; groups vulnerable to disinformation such as
the elderly, young people, etc.); society in general; and other. ere was also a eld open to comments.
A short questionnaire was chosen, mostly of closed questions and with some non-mandatory responses, as well as space for
comments. It consisted of several sections, one generic and others specic to the aspects highlighted (use of AI, development
of proprietary AI tools, and training/literacy initiatives on AI and disinformation), with a sequence of questions that allowed
participants to submit the questionnaire without having to answer all the questions if they had not used AI in any way or
had responded negatively to subsequent questions about specic uses. Instructions were added to the nal online design via
Google Forms, to thus facilitate responses in several languages, while an invitation letter was added in English and Spanish.
at method of asynchronous, written response was considered more appropriate given the nature of the projects analysed,
which spanned dierent countries and languages.
It is also worth noting that two reminders were sent, to ensure a high response rate, and direct contact was made, in some
cases, through professional social media such as LinkedIN.
doxa.comunicación |
nº 41, pp. 489-509
July-December of 2025
Ana Zafra Arroyo, María Sánchez González and Hada M. Sánchez Gonzales
ISSN: 1696-019X / e-ISSN: 2386-3978
| 499
A total of 14 responses were obtained to the online questionnaire between September 3 and 24, 2024, representing just under
half of the projects surveyed. e following table summarises the participants’ data.
Table 2. Projects analysed from which a response was obtained
Project name
Name of the person
answering the questionnaire
Responsibility/position in the project
Escape Fake 2.0Irina ParaschivoiuProject Lead
Made for social media: Fighting disinformation with
fact check graphics
Dominika Chronowska
Journalist
Czech Elections in the Era of Disinformation: e 2023
Presidential Election
Kristína
Šef
č
íková
Project Manager
Empowering Schools in self- regulation of Media and
Information Literacy
Christian Tarchi
Coordinator
Disinformation LaundromatHelena SchwertheimProject Manager
Stars4Media-FACTCHECKING
ierry LeroyProject coordinator
Ukrainian news Agency to integrate fact-checked
content with the help of AFP and APA
Florian Schmidt
Trainer
YO-MEDIA: Youngsters’ Media Literacy in Times of
Crisis
Nadine SantosOperations Project Manager
Scale-up Science Fact-Checking in Spanish Podcasts
and Youtube
Alba Tobella
Director
Open Source Munitions PortalGiacomo NanniProject Lead
FactSphere: Exposing Disinformation for Public
Awareness in the Black Sea Region
Voinea MadalinaDisinformation Specialist
Boosting Fact-checking Activities in EuropeLaura Burtan
Researcher
TeaMLit
Sally Reynolds
Coordinator
UDDOT: Understanding Disinformation Dynamics
on Telegram
Pablo Hernández
Academic research coordinator at
Maldita.es
, project leader
Source: created by the authors
<p>[Enlace de URL / hc (has AS)]</p>
500 |
nº 41, pp. 489-509
|
July-December of 2025
Innovations using Generative AI for media literacy and fact-checking in the European Union
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Given the study’s exploratory nature, most of the questions were closed-ended, designed to identify general trends in the
projects. An inductive thematic analysis was applied to the open- ended questions, a common technique in qualitative
research that allows for the identication of patterns of meaning from textual content (Braun & Clarke, 2006). e responses
were read by the research team and manually coded to identify recurring themes.
3. Results
14 of the 39 projects funded by the EMIF have taken part in the questionnaire as of the date of this paper, representing a
response rate of 35.9%. 10 of those projects—equivalent to 25.6% of the total number of funded projects—reported having
incorporated some type of innovation, either through methodological approaches such as gamication or through the use of
advanced technological resources. Furthermore, six of the responding projects mentioned having used GenAI at some stage in
their development. However, only ve of the cases (12.8% of the total of funded projects) were considered valid for the analysis,
as one of them corresponded to a dierent call and did not meet the inclusion criteria dened in this study.
Although the primary objective of this study was to analyse the presence of GenAI in EMIF- funded projects, the results also
allowed identication of other relevant emerging trends, particularly regarding the adoption of innovative technologies such
as gamication in media literacy processes and automation applied to fact-checking.
3.1. Technological innovation and the use of AI in media literacy
Media literacy is a way to combat disinformation. e discipline has integrated innovative new technologies oriented towards
gamication and the fostering of more interactive teaching. e EMILE project, headed by the University of Florence, uses
a video game called ‘Elli’s World’ which employs data mining to adapt the game’s content and diculty to each student’s
ability. at allows identication of cognitive proles and the tailoring of the learning experience to suit individual needs.
‘Elli’s World’ recreates a simulated digital environment and encourages skills such as critical analysis and informed decision-
making. Furthermore, an option for teachers allows the real time monitoring of the student’s progress and to step in and adapt
teaching as necessary.
e Escape Fake 2.0 project is a digital escape room-style game which is aimed at instilling fake news detection skills. It allows
students to confront situations of disinformation in an environment that simulates social media. Participants have to decipher
codes, verify the authenticity of content, and collect evidence. ey thus hone their ability to distinguish between true and
false information. Furthermore, the game incorporates a pedagogical agent that provides clues and suggestions. at can
guide players and prevent frustration while they perform tasks such as reverse image searches and identifying reliable sources.
According to information provided by the project’s representative, Escape Fake 2.0 has used AI tools to generate text, image,
and video content for integration with educational play. e organisation has internal guidelines regulating the use of AI and
uses paid-for versions of Midjourney and Chat GPT, which gives them the assurance that companies cannot train their models
based on data from Escape Fake 2.0.
Although gamication through the use of technologies such as VR or AR is the most widespread innovative approach in the
projects studied, YO-MEDIA has chosen to include physical experiences in addition to digital ones. e objective is the same
doxa.comunicación |
nº 41, pp. 489-509
July-December of 2025
Ana Zafra Arroyo, María Sánchez González and Hada M. Sánchez Gonzales
ISSN: 1696-019X / e-ISSN: 2386-3978
| 501
as that of the games mentioned above: to improve young students’ ability to identify fake news and their critical analysis in
crisis contexts (health emergencies, armed conicts, natural disasters, etc.). YO- MEDIA recreates high-pressure scenarios
where participants need to discern between trustworthy information and disinformation. e project also oers Massive Open
Online Courses (MOOCs), which reinforce the practical skills acquired through theorizing concepts.
Finally, TeaMLit focuses on preparing teachers in media and information literacy (MIL) skills. e goal is to integrate these
skills into the school curriculum. e project provides teachers with online resources and databases that make it easier to
teach MIL at various educational levels.
3.2. Technological innovation and use of AI in fact-checking
Among the projects that carry out fact-checking activities, technological innovations have been discovered aimed at making
it faster and easier to detect disinformation, including the use of predictive or generative articial intelligence. An example
is SCENT, led by Science Feedback, which uses a network of backlinks or hyperlinks to assess the credibility of websites by
analysing link patterns between web domains. e project employs a Relational Graph Convolutional Network (RGCN) model,
a deep learning technique that allows for the processing of large volumes of structural data. After analysing more than 700,000
domains, SCENT has processed a considerable number of networks of both information and disinformation. Just as high
standard domains tend to link to each other, low-credibility sites also tend to link to similar websites of questionable data
quality. SCENT automates the evaluation of the credibility of scientic websites, providing both fact-checkers and the general
public with access to a public database where they can query the credibility of a domain in real time by means of an Application
Programming Interface (API).
e FAST system, also aimed at improving fact-checking processing times, has been developed by
Maldita.es
and Servimedia.
It is based on the early detection of emerging disinformation campaigns using an algorithm capable of recognising them. us,
FAST focuses not only on debunking (refuting existing disinformation) but also on prebunking (pre- emptive refutation), a
dual strategy that allows the system to operate proactively, mitigating the impact of disinformation before it reaches a mass
audience.
Another tool designed to facilitate fact-checking procedures is Scale-up Science Fact Checking, developed by the Spanish
fact-checking platform Vericat. Scale-up performs real-time analysis to verify audiovisual content. e system uses AI and
voice recognition technologies to examine large volumes of content on platforms such as YouTube as well as in podcasts. Fact-
checkers can utilise the tool to automatically transcribe up to 5,000 hours of video and 2,500 hours of audio, enabling the study
of disinformation patterns on topics such as health and science. A dashboard that organises the transcribed content allows for
the ecient identication of misleading narratives. Scale-up’s representative, in response to the questionnaire, specied that
they are working jointly with Chequeado, an Argentine fact- checking company, to jointly use Chequeabot, the microsite-type
tool developed by the Argentine rm. e tool is used solely within the framework of the project to detect veriable statements
in podcasts and YouTube by use of keywords. However, future funding for the use of Chequeabot is not guaranteed, according
to Scale-up’s representative. She also mentions the diculty of nding AI tools that work in less widely-spoken languages such
as Catalan.
<p>[Enlace de URL / hc (has AS)]</p>
502 |
nº 41, pp. 489-509
|
July-December of 2025
Innovations using Generative AI for media literacy and fact-checking in the European Union
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
e Disinformation Laundromat, a project led by the Institute for Strategic Dialogue (ISD) and the University of Amsterdam,
has developed a tool that enables the automated detection of disinformation ampliers by identifying their funding sources.
is is an Open-Source Intelligence (OSINT) technology. e Disinformation Laundromat uses algorithms to automate the
detection of malicious websites, drawing up a map of distribution networks and their associated channels of monetisation.
By bringing together several OSINT sources into a single platform, the project enables researchers to carry out an integrated
analysis of the links between malicious actors and their content networks.
Although they have not developed their own tool, the
Ukrainian News Agency to Integrate Fact-Checked Content with the Help
of AFP and APA
project have integrated AI into their fact- checking routines as a resource for comparing multiple types of
content, including images, videos, and previously published quotes, taking advantage of various freely-available fact- checking
tools. Similarly, the members of Made for Social Media, whose goal is to combat disinformation through fact-checked graphics
on social media, use free AI tools to verify images and validate sources.
3.3. Perceptions of GenAI and future trends in its use in combating disinformation
e questionnaire focused on the projects included in the 2022/2023 report and those funded as of June 2024. It allows for
the identication of general trends, but its scope is limited regarding the use of GenAI, as the technology was not yet widely
implanted during that period. Some predominant trends or areas of use have been observed among those that do use it. ose
include streamlining tasks related to project activity. Other respondents report using AI to fact check and compare sources.
Finally, the development of proprietary AI tools or collaboration on their development with other organizations is relatively
rare.
ere is consensus in the responses studied regarding the statements “AI makes it easier for fact-checkers and news
professionals to verify information” and “AI is a tool that allows citizens to check facts for themselves.” However, there are
ambivalent positions regarding the idea that AI does not yet have an impact on journalism. at is, on the Likert scale of 1 to 3,
with 3 showing the highest degree of agreement, the majority responded with a 2, with 3 being the next most voted response.
Nor are clear positions or trends apparent regarding whether AI has an impact, beyond the ght against disinformation, on
information manipulation and the spread of fake news.
doxa.comunicación |
nº 41, pp. 489-509
July-December of 2025
Ana Zafra Arroyo, María Sánchez González and Hada M. Sánchez Gonzales
ISSN: 1696-019X / e-ISSN: 2386-3978
| 503
Graph 1. Representation of absolute frequencies of responses on a Likert scale by statement
Source: created by the authors
e questionnaire also posed the question: “In general terms, what possibilities or potential do you see, based on your
knowledge and experience, for Generative Articial Intelligence to serve in the ght against disinformation?” e aim thereby
was to identify the main trends in the application of this tool to mitigate information disorders. Practically all the responses
mentioned the possibility of analysing social media and large volumes of data. Among these was Scape Fake, which points
out that detecting disinformation spread patterns on social media using GenAI can support the work of journalists and fact-
checkers. Projects such as e Disinformation Laundromat, Fact Sphere, and Made for Social Media also recognise that
the potential of GenAI lies in its ability to be trained on large datasets. Once the technology is trained “with disinformation
patterns,” it can be utilised to automatically detect false narratives.
Reference is also made to the integration of this learning into newer formats such as memes and short videos. e Disinformation
Laundromat mentions that the analysis of non-textual formats such as images and videos is a major advantage, as is the speed
with which it can be performed. According to the project’s representative, systems that integrate GenAI “could monitor and
ag misleading information more quickly and eectively than human moderators.” Open Source Munitions Portal states that
algorithms for cross-referencing material and analysing images also allow for systematic content verication. at is, the
cross- checking of sources or images from dierent platforms to determine their authenticity.
Source:
created by the
authors
The
questionnaire
also
posed
the
question:
"In
general
terms,
what
possibilities
or
potential
do
you see, based on your knowledge and experience, for Generative Artificial Intelligence to
serve in the fight against disinformation?" The aim thereby was to identify the main trends in
the application of this tool to mitigate information disorders. Practically all the responses
mentioned the possibility of analysing social media and large volumes of data. Among these
was
Scape
Fake,
which
points
out
that
detecting
disinformation
spread
patterns
on
social
media
using GenAI can support the work of journalists and fact
-
checkers. Projects such as The
Disinformation Laundromat, Fact Sphere, and Made for Social Media also recognise that the
potential of GenAI lies in its ability to be trained on large datasets. Once the technology is
trained
"with
disinformation
patterns,"
it
can
be
utilised
to
automatically
detect
false
narratives.
Reference is also made to the integration of this learning into newer formats such as memes
and short videos. The Disinformation Laundromat mentions that the analysis of non
-
textual
formats such as images and videos is a major advantage, as is the speed with which it can be
performed. According to the project’s representative, systems that integrate GenAI "could
monitor and flag misleading information more quickly and effectively than human
moderators." Open Source Munitions Portal states that algorithms for cross
-
referencing
material
and
analysing
images
also
allow
for
systematic
content
verification.
That
is,
the
cross-
checking of sources or images from different platforms to determine their authenticity.
Scale-
up Science Fact
-
Checking is an example of how to apply GenAI to analyse audio and
text to identify misinformative content. The project maintains that beyond this functionality,
AI-
powered content generation can be employed to moderate comments and messages in
virtual
communities.
They
point
out
that
it
may
be
possible
to
create
a
chatbot
on
social
media
which can be employed for clarification by users who encounter content of questionable
veracity. Similarly, other projects highlight GenAI which can design proactive responses to
counter disinformation.
Please
indicate,
on
a
scale
of
0
to
3,
the
degree
to
which
you
agree
with
each
of
the
following
statements:
La
IA
no
tiene
aún
una
verdadera
incidencia
en
el
periodismo
La IA incide, más que en la lucha contra la
desinformación,
en
la
manipulación
informativa
y
en la expansión de fake news.
La
IA
es
una
herramienta
para
que
la
ciudadanía
pueda por sí misma verificar información
La
IA
facilita
la
verificación
de
información
a
fact-checkers
y
profesionales
de
lainformación
504 |
nº 41, pp. 489-509
|
July-December of 2025
Innovations using Generative AI for media literacy and fact-checking in the European Union
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Scale-up Science Fact-Checking is an example of how to apply GenAI to analyse audio and text to identify misinformative
content. e project maintains that beyond this functionality, AI-powered content generation can be employed to moderate
comments and messages in virtual communities. ey point out that it may be possible to create a chatbot on social media
which can be employed for clarication by users who encounter content of questionable veracity. Similarly, other projects
highlight GenAI which can design proactive responses to counter disinformation.
is is the case of the
Ukrainian News Agency to integrate fact-checked content with the help of AFP and APA
, which argues that
such a tool would be useful in distributing trustworthy
content. inking especially of young people, they argue that AI tools –whether generative or not– have great potential in
disseminating fact-checks. e project’s representative suggests that fact-checks are more engaging in audiovisual formats, but
in his opinion, that is not a simple task. He asserts that the use of AI facilitates the distribution of veried narratives that do not
consist solely of vast texts that are of little interest to most audiences.
In the educational eld, Empowering Schools in Self-Regulation of Media and Information Literacy views GenAI as an
innovation that can help students understand complex texts by simplifying them. Furthermore, it can provide feedback on
the critical analysis that young people make of the information available to them. e general trend is that if GenAI continues
to evolve in terms of advanced analysis, fact-checking, and data generation processes, “it will allow texts to be analysed and
readers to be informed in real time when they encounter false information,” as proposed by Made for Social Media.
4. Discussion
is analysis of projects supported by the European Media and Information Fund (EMIF) provides a snapshot of innovations,
involving the use of generative AI, to combat disinformation and to educate and prepare society against hoaxes. e projects
studied stand out for their ability to automatically detect false information, sift through large volumes of information on social
media, and generate responses. e incorporation of AI into formats such as memes and short videos is also pointed to as a
signicant advantage. Such is the case with e Disinformation Laundromat and Made for Social Media who highlight the
eciency and agility with which AI can detect misleading information and perform in-depth cross-checking of dierent
platforms.
Literacy and training are essential in the ght against disinformation, as other studies have already propounded in the Ibero-
American context (Sánchez & Sánchez, 2022) and in Europe (Sádaba & Salaverría, 2023; Sánchez et al, 2024). is study has
revealed initiatives that foster interactive teaching to eectively address the challenges of disinformation through gamication
in AI-generated virtual environments as well as the physical world. us, they not only improve learners’ skills but also help
educators eectively address the challenges of disinformation.
While AI certainly does oer several benets, it also entails challenges and risks of information manipulation and false
content. erefore, the introduction of eective regulatory measures and guidelines is essential, as highlighted in other studies
(Bontridder & Poullet, 2021; Sánchez, 2022). at is the task of SOLARIS, a European project that analyses the dangers and
establishes regulations to mitigate the risks of disinformation (UC3M, 2023). Reliance on GenAI may lead to a lack of thinking
doxa.comunicación |
nº 41, pp. 489-509
July-December of 2025
Ana Zafra Arroyo, María Sánchez González and Hada M. Sánchez Gonzales
ISSN: 1696-019X / e-ISSN: 2386-3978
| 505
and skills in fact-checking. erefore, AI literacy and training are essential in preparing students and society in general for the
challenges of the digital environment.
e perception among fact-checkers is that the potential of AI is a positive and its future signicant, but it requires proper
management and policies to address the challenges of manipulation. e connection between organisations, as previously
highlighted by other authors (Sánchez, Sánchez & Martos, 2024), the regulation of tools, and the advancement of studies in
this eld will be essential to address the challenges of disinformation related to the use of GenAI due to its constant evolution
in data analysis and creation, as well as in comment moderation.
5. Conclusions
Over two years have now passed since Open AI’s launch of Chat GPT, and the landscape of perception and use of generative
articial intelligence by media companies is advancing by leaps and bounds. While much remains to be done, one could
say that we have moved on from initial fears or scepticism to beginning to employ this technology for certain issues and
even launching training initiatives for newsroom sta or starting to launch some regulations or white papers to ensure proper
standards in the use of GenAI.
Regarding the ght against disinformation, some fact-checking platforms in both Latin America and Europe have been
leading, due to their nature as startups and their orientation towards technological innovation, certain projects to develop
applications and AI tools for the ght against disinformation and have been pioneers in training on this subject (Sánchez,
Sánchez & Martínez, 2022 and Sánchez, Sánchez & Alonso, 2024).
e projects analysed share with those fact-checking platforms the use of AI in visual content to quickly and eectively recognise
patterns of disinformation, detect false content, and analyse and monitor information from social media, as previously
indicated (Sánchez, Sánchez, & Martos, 2024). However, this study reveals the application of technological innovations for
literacy and training in the use of GenAI focused on gamication and automation so that students can eectively identify false
information and personalise their learning. at is the case of the projects EMILE, Escape Fake 2.0, and YO-MEDIA.
Another highlight is the use of deep learning techniques through Relational Graph Convolutional Networks (RGCN). ey
have been utilised by projects such as SCENT, FAST, and Scale-up Science Fact Checking to detect disinformation campaigns
and assess website credibility, as well as for automated, real-time detection of fake content (e Disinformation Laundromat
and Made for Social Media).
us, there is a consensus among project leaders on the importance and potential of GenAI to facilitate fact-checking and help
mitigate disinformation. ey believe it is essential to integrate literacy skills into the school curriculum, as the TeaMLit project
leader emphasized.
As AI evolves and signicant advances are made, there will also be a need to move forward with policies regulating the use
of these tools, as well as with education, fact-checking, and research into the topic that sheds light on how to better combat
disinformation.
506 |
nº 41, pp. 489-509
|
July-December of 2025
Innovations using Generative AI for media literacy and fact-checking in the European Union
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
is study oers two important contributions. Firstly, its pioneering nature, as it addresses a still-nascent eld, the integration
of GenAI into media literacy projects. Secondly, it focuses on initiatives using European public funds, which represents
an exercise in social responsibility and accountability by analysing how public resources are invested in the ght against
disinformation.
However, precisely because of the period in which it was conducted—focusing primarily on projects included in the 2022/2023
European Media and Information Fund (EMIF) report—its results are necessarily limited in terms of detecting the use of
GenAI, a technology that had not yet achieved widespread implementation at that time. Although several important projects
and trends were identied, both the adoption of the technology and mentions of its use in the questionnaire responses were
few, which reduced the sample ultimately analysable in this specic dimension.
6. Acknowledgments
is article has been translated into English by Brian O”Halloran to whom we are grateful for his work.
Excellence project, Andalusian Plan for Research, Development and Innovation (PAIDI 2020): “e impact of disinformation
in Andalusia: cross-sectional analysis of audiences and journalists’ routines and agendas. Disinfoand Ref: ProyExc00143.
2022-2025
7. Specic contributions of each author
Name and Surname
Conception and design of the work
Ana Zafra Arroyo, María Sánchez González and
Hada M. Sánchez Gonzales
Methodology
María Sánchez González
Data collection and analysisAna Zafra Arroyo and María Sánchez González
Discussion and conclusionsHada M. Sánchez Gonzales
Drafting, formatting, version review and approval
Ana Zafra Arroyo, María Sánchez González and
Hada M. Sánchez Gonzales
8. Conicts of interest
e authors declare that there is no conict of interest contained in this article.
doxa.comunicación |
nº 41, pp. 489-509
July-December of 2025
Ana Zafra Arroyo, María Sánchez González and Hada M. Sánchez Gonzales
ISSN: 1696-019X / e-ISSN: 2386-3978
| 507
9. Bibliographical references
Allyn, B. (11 de octubre de 2024). TikTok executives know about app’s eect on teens, lawsuit documents allege.
NPR
.
bit.
ly/3Fxz48A
Adair, B. (23 de febrero de 2020).
Squash report card: Improvements during State of the Union ... and how humans will make our
AI smarter.
Duke Reporters’ Lab
.
https://bit.ly/4kKSnu5
Bharti, I., Chauhan, K., & Aggarwal, P. (2025). Generative AI: Next frontier for competitive advantage. En A. Natarajan, M.
Galety, C. Iwendi, D. Das, & A. Shankar (Eds.),
Enhancing communication and decision-making with AI
(pp. 1-36). IGI Global.
https://doi.org/10.4018/979-8-3693-9246-1.ch001
Black, J. (2008). Constructing and contesting legitimacy and accountability in polycentric regulatory regimes.
Regulation &
Governance, 2
(2), 137–164.
https://doi.org/10.1111/j.1748-5991.2008.00034.x
Bontridder, N., & Poullet, Y. (2021). e role of articial intelligence in disinformation.
Data & Policy, 3
(32).
https://doi.org/10.1017/dap.2021.20
Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology.
Qualitative Research in Psychology, 3
(2), 77–101.
https://doi.org/10.1191/1478088706qp063oa
Calouste Gulbenkian Foundation. (s. f.).
European Media and Information Fund at a glance
. European Media and Information
Fund. Recuperado el 19 de diciembre de 2024, de
https://gulbenkian.pt/emifund/emif-at-a-glance/
Calouste Gulbenkian Foundation. (s. f.).
FAST: Fact-checkers acting to stop early-detected disinformation campaigns
. European
Media and Information Fund. Recuperado el 19 de diciembre de 2024, de
https://bit.ly/43IJavD
Calouste Gulbenkian Foundation. (s. f.).
e disinformation laundromat: An OSINT tool to detect disinformation and amplication
networks
. European Media and Information Fund. Recuperado el 19 de diciembre de 2024, de
https://bit.ly/4kj3XN6
de Cock Buning, M. (2018).
A multi-dimensional approach to disinformation: Report of the independent high level group on fake
news and online disinformation
. Publications Oce of the European Union.
European Commission. Directorate-General for Communication Networks, Content, & Technology. (2018).
A multi-dimensional
approach to disinformation : Report of the independent High Level group on fake news and online disinformation.
Publications
Oce of the European Union.
https://data.europa.eu/doi/10.2759/739290
European media and information fund. (4 de octubre de 2021). European Media and Information Fund.
https://gulbenkian.pt/
emifund/
Gamito, M. C. (2023). e European Media Freedom Act (EMFA) as meta-regulation.
Computer Law & Security Review, 48
, 105799.
https://doi.org/10.1016/j.clsr.2023.105799
Garriga, M., Ruiz-Incertis , R., & Magallón-Rosa, R. (2024). Articial intelligence, disinformation and media literacy proposals
around deepfakes.
Observatorio (OBS*)
, 18(5).
https://doi.org/10.15847/obsOBS18520242445
Gonzalo, M. (8 de febrero de 2023). El cierre de la API de Twitter amenaza la lucha contra la desinformación, el acoso y los
contenidos tóxicos.
Newtral
.
https://bit.ly/3HvMna8
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
508 |
nº 41, pp. 489-509
|
July-December of 2025
Innovations using Generative AI for media literacy and fact-checking in the European Union
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Gutiérrez Caneda, B., & Vázquez-Herrero, J. (2024). Redibujando las líneas contra la desinformación: Cómo la IA está modicando
el presente y futuro del fact-checking.
Tripodos,
(55), 55–74.
https://doi.org/10.51698/tripodos.2024.55.04
Haggart, B., Tusikov, N., & Scholte, J. A. (Eds.). (2021).
Power and Authority in Internet Governance: Return of the State?
(1st ed.).
Routledge.
https://doi.org/10.4324/9781003008309
Herke, C. (2023). Deepfake.
Pro Futuro.
https://doi.org/10.26521/profuturo/2023/1/13334
Hetler, A. (2024, julio 31).
What is ChatGPT? Everything you need to know.
WhatIs; TechTarget. https://www.techtarget.com/
whatis/denition/ChatGPT
Lampropoulos, G., & Kinshuk. (2024).
Virtual reality and gamication in education: A systematic review. Educational Technology
Research and Development, 72,
1691–1785.
https://doi.org/10.1007/s11423-024-10351-3
Machiraju, S., Modi, R. (2018). Natural Language Processing. En Developing Bots with Microsoft Bots Framework. Apress,
Berkeley, CA.
https://doi.org/10.1007/978-1-4842-3312-2_9
Maldita.es
. (2023, 8 de febrero).
Maldita.es
y Servimedia recibirán una beca del EMIF para detectar y combatir la desinformación
.
Recuperado de
https://bit.ly/43ju4xy
Miranda, S., Vlachos, A., Nogueira, D., Secker, A., Mendes, A., Garrett, R., Mitchell, J. J., & Marinho, Z. (2019). Automated fact
checking in the news room.
e Web Conference 2019: Proceedings of the World Wide Web Conference, WWW 2019,
3579-3583.
Association for Computing Machinery.
https://doi.org/10.1145/3308558.3314135
Montoro-Montarroso, A., Cantón-Correa, J., Rosso, P., Chulvi, B., Panizo-Lledot, Á., Huertas-Tato, J., Calvo-Figueras, B.,
Rementeria, M. J., & Gómez-Romero, J. (2023). Fighting disinformation with articial intelligence: Fundamentals, advances and
challenges.
Profesional de la Información, 32(
3).
https://doi.org/10.3145/epi.2023.may.22
Mungiu-Pippidi, A. (2015).
e quest for good governance: How societies develop control of corruption.
Cambridge University
Press.
https://doi.org/10.1017/CBO9781316286937
Park, K., & Culloty, E. (2023). Beyond performative transparency: lessons learned from the EU Code of Practice on Disinformation.
AoIR Selected Papers of Internet Research,
2022.
https://doi.org/10.5210/spir.v2022i0.13067
Pecini, C., Della Rocca, S., y Scrocco, A. (2024). Critical Reading of Digital Texts: e Emile Project Empowering Schools in Self-
Regulation of Media and Information Literacy Processes.
https://bit.ly/43niv8w
Prodigioso Volcán (2023).
IA para periodistas. Una herramienta por explotar.
https://bit.ly/4kHhopW
Rzabayeva, D., Kassymova, G., & Pratama, H. (2024).
e Role of Gamication in Promoting Digital Literacy: Bridging the
Gap between Fun and Learning. Materials of International Scientic-Practical Internet Conference “Challenges of Science”.
https://doi.org/10.31643/2024.20
Sádaba, C. & Salaverría, R. (2023). Combatir la desinformación con alfabetización mediática: análisis de las tendencias en la
Unión Europea.
Revista Latina de Comunicación Social, 81
, 1–17.
https://doi.org/10.4185/RLCS-2023-1552
Sánchez, H. M. (2022). Transformación digital y audiencia. Tendencias y uso de la inteligencia articial en medios vericadores.
Ámbitos
, 56, 9-20.
https://doi.org/10.12795/Ambitos.2022.i56.01
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
doxa.comunicación |
nº 41, pp. 489-509
July-December of 2025
Ana Zafra Arroyo, María Sánchez González and Hada M. Sánchez Gonzales
ISSN: 1696-019X / e-ISSN: 2386-3978
| 509
Sánchez, H. M. ; Sánchez, M, & Martos, J. (2024). e methodology used by fact-checkers. An in-depth analysis of commonly used
strategies.
Journalism Practice
, 1-24,
https://doi.org/10.1080/17512786.2024.2340522
Sánchez, M. & Sánchez, H. M., (2022). Fact-checkers como formadores: iniciativas de los iberoamericanos de IFCN en Escandón,
P. Tejedor, S. (ed.).
Escenarios digitales de la comunicación
(pp.19-36). Gedisa.
Sánchez, M., Sánchez, H. M., & Martínez, S. (2022). Inteligencia articial en vericadores hispanos de la red IFCN:
proyectos innovadores y percepción de expertos y profesionales.
Estudios sobre el Mensaje Periodístico
, 28(4), 867-879.
https://doi.org/10.5209/esmp.82735
Santos, F. C. C. (2023). Articial Intelligence in Automated Detection of Disinformation
: A ematic Analysis. Journalism and
Media, 4
(2), 679-687.
https://doi.org/10.3390/journalmedia4020043
Scharrer, E., & Zhou, Y. (2022, May 18). Media Literacy and Communication.
Oxford Research Encyclopedia of Communication.
https://doi.org/10.1093/acrefore/9780190228613.013.1304
Science Feedback. (s. f.).
SCENT: Predicting domain credibility on the basis of backlink networks
. Recuperado el 19 de diciembre
de 2024, de
https://bit.ly/3Fg1QdZ
Scott, J. C. (2018).
Lobbying and society: A political sociology of interest groups.
Polity.
Sedelmeier, U. (2012). Is Europeanisation through conditionality sustainable? Lock-in of institutional change after EU accession.
West European Politics, 35
(1), 20-38.
https://doi.org/10.1080/01402382.2012.631311
Singh, J. N., Gautam, A., & Tomar, H. (2023). Deep Fake in picture using Convolutional Neural Network. En
Proceedings of the
2023 5th International Conference on Advances in Computing, Communication Control and Networking (ICAC3N).
IEEE.
https://doi.org/10.1109/ICAC3N60023.2023.10541758
e Information Laundromat. (s. f.).
e Information Laundromat
. Recuperado el 19 de diciembre de 2024, de
https://informationlaundromat.com/
Tiernan, P., Costello, E., Donlon, E., Parysz, M., & Scriney, M. (2023). Information and media literacy in the age of AI: Options for
the future.
Education Sciences, 13
(9), 906. https://doi.org/10.3390/educsci13090906
UC3M. (2023).
Un proyecto cientíco europeo estudia cómo regular la desinformación generada por IA
. Universidad Carlos III de
Madrid.
Waterson, J. (2019, mayo 24). Facebook refuses to delete fake Pelosi video spread by Trump supporters.
e guardian.
https://bit.ly/4kNGgwx
Wetzel, D. A., & Kani, J. (2024). Enhancing information literacy through generative AI in the library classroom. Pennsylvania
Libraries:
Research & Practice, 12(
2).
https://doi.org/10.5195/palrap.2024.302
Woolley, J. (2024). Generative AI and Business: A Review and Research Agenda. Oxford Research Encyclopedia Of Business And
Management.
https://doi.org/10.1093/acrefore/9780190224851.013.434
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>