51.5 per cent of the Spanish population feel capable of spotting misinformation about science, but only 18.1 per cent believe others are capable of doing so

There are several findings in the report that are fully supported by the data collected and are worth highlighting, including the increase in the use of AI as a source of information, people’s tendency to interpret science based on certain assumptions — as in the case of scientific populism —, the importance of reflecting before sharing content, and so on. These are precisely the conclusions that have caught my attention most in this report, which I consider a significant contribution for those of us who work daily in science communication research.

The report provides evidence on current lines of research in the field of science communication, particularly regarding how the public accesses information passively (News Finds Me), or the use of social media and artificial intelligence to stay informed (the platformisation of content). Furthermore, it is in line with all international research on scientific literacy (formal, non-formal and informal science) and on the emotional impact, beliefs, ideology and values that prevail when consuming and sharing scientific information and disinformation (public attitudes towards science).

The limitations of the study are highlighted by the authors in relation to the formulation of specific questions and also regarding the restrictive nature of the topics covered. In this regard, I believe the proposal is methodologically well-conceived and consistent with the stated objectives. However, I believe that in subsequent studies a distinction could be made between sources of information (people: politicians, experts, activists, influencers, journalists, etc.; teams, studies, communities, documents) and channels (platforms, channels, search engines, outdoor advertising, etc.); and that a focus group could be incorporated to identify arguments that might complement the quantitative data.

On trust in science

In my view, the most significant finding of the report is that the public continues to place a great deal of trust in researchers, and this is of great value in a context such as the current one, which is clearly characterised by information overload and the growing spread of disinformation. However, seeing once again from the results that this trust in science suffers when scientific knowledge is transferred to the realm of political institutions, governance or certain intermediaries, such as the media, struck me as an area of opportunity for science communication.

I believe that one of the main keys to interpreting these results lies precisely in that gap between the trust that science inspires and the mistrust towards the system responsible for communicating it, managing it or translating it into public policy. Therefore, bridging that gap constitutes, in my view, one of the major challenges facing science communication today.

On scientific populism

Furthermore, the report shows that the spread of scientific disinformation does not depend exclusively on the existence of false content in the public sphere, but is closely linked to people’s deeper predispositions, such as magical thinking, conspiracy theories or scientific populism,

that is, the tendency to prioritise personal experience over knowledge based on available evidence (as in the example: ‘it has worked for me or for people I know’). This would suggest that, in order to combat disinformation, in addition to identifying hoaxes, we must take into account the interpretative frameworks that lead certain people to regard them as credible. Hence the importance of all the qualitative studies carried out in the field of science communication research to understand the public’s frames of perception and interpretation.

On thinking before sharing

One of the findings I found most encouraging in the report is the evidence that pausing to reflect (or stopping to think) before sharing information you receive helps to minimise the spread of false content. This highlights that when people pause to ask themselves whether a news item is credible or whether it is worth verifying, their intention to spread false content decreases significantly. In other words, they act less impulsively. The experiment carried out for the report shows that small interventions that encourage critical reflection can be very effective in reducing the circulation of disinformation.

On scientific and media literacy

I believe the report also offers a very important lesson for public policy, namely that having more years of formal education does not, in itself, protect against scientific disinformation. Instead, what really makes the difference is understanding how scientific knowledge works and how the current information ecosystem operates. For this reason, scientific literacy and media literacy are establishing themselves as fundamental tools for strengthening social resilience against disinformation. In this regard, the European Digital Media Observatory is committed, through all the hubs that make up the Observatory, to promoting media literacy activities to empower citizens and improve resilience against disinformation.

On AI as a source of information

One of the most significant trends highlighted in this report is the speed with which artificial intelligence is becoming a common source of scientific information, particularly amongst younger people. In my view, the real turning point is not merely that more and more people are consulting these systems, but that they tend to perceive them as objective, neutral and autonomous technologies, when in reality they are not.

We are currently working with an international group on the biases present in the information generated by artificial intelligence in the field of science, and we start from the premise that AI does not produce knowledge in isolation from society, as it learns from data generated by people, is developed in accordance with specific technical and business decisions, and responds to the specific values, priorities and interests of major technology developers. For this reason, it can amplify biases, render certain knowledge invisible or reinforce erroneous narratives. If AI is to become a new gateway to scientific information, we should demand transparency regarding the sources it uses, the criteria by which it prioritises information, and the mechanisms by which it manages uncertainty and error. The question is: who decides how AI learns, what it shows us and what it leaves out when it comes to scientific knowledge?

Sobre populismo científico 

Por otra parte, el informe muestra que la difusión de la desinformación científica no depende exclusivamente de la existencia de contenidos falsos en la esfera pública, sino que está muy relacionada con predisposiciones más profundas de las personas, como, por ejemplo, con el pensamiento mágico, conspirativo o con el populismo científico, es decir, con esa tendencia a utilizar la experiencia personal por encima del conocimiento basado en la evidencia disponible (el ejemplo de “a mí o mis personas conocidas les ha funcionado”). Esto significaría que, para combatir la desinformación, además del trabajo de identificación de los bulos, habría que tener en cuenta qué marcos de interpretación llevan a determinadas personas a considerarlos creíbles. De ahí la importancia de todos los estudios cualitativos que se llevan en el ámbito de la investigación de la comunicación de la ciencia para conocer los marcos de percepción e interpretación de la ciudadanía. 

Sobre reflexionar antes de compartir 

Uno de los resultados que me ha parecido más positivo del informe es comprobar que introducir una pausa (una reflexión o pararse a pensar), antes de compartir la información que te llega, funciona para minimizar la diseminación de contenidos falsos. Esto pone de manifiesto que cuando las personas se detienen a preguntarse si una noticia es creíble o si merece la pena verificarla, disminuye significativamente su intención de difundir contenidos falsos. Es decir, se actúa de forma menos impulsiva. El experimento que se ha realizado para la elaboración del informe evidencia que pequeñas intervenciones que fomenten la reflexión crítica pueden ser muy eficaces para reducir la circulación de la desinformación. 

Sobre alfabetización científica o mediática 

Creo que el informe también deja una enseñanza muy importante para las políticas públicas y es que tener más años de formación reglada no protegen, por sí solos, frente a la desinformación científica. En cambio, lo que realmente marca la diferencia es comprender cómo funciona el conocimiento científico y cómo opera el ecosistema informativo actual. Por ello, la alfabetización científica y la alfabetización mediática se consolidan como herramientas fundamentales para fortalecer la resiliencia social frente a la desinformación. En este sentido, el European Digital Media Observatory está apostando desde todos los hubs que configuran el Observatorio por promover actividades de alfabetización mediática para el empoderamiento ciudadano y mejorar la resiliencia frente a la desinformación.  

Sobre la IA como fuente de información 

Uno de los cambios más significativos en cuanto a tendencias que se recoge en este informe es la rapidez con la que la inteligencia artificial está convirtiéndose en una fuente habitual de información científica, especialmente entre la población más joven. Desde mi punto de vista, el verdadero punto de inflexión no es solo que cada vez más personas consulten estos sistemas, sino que tienden a percibirlos como tecnologías objetivas, neutrales y autónomas, cuando en realidad no lo son. 

Estamos trabajando justo en este momento con un grupo internacional, sobre los sesgos que se producen en la información que arroja la inteligencia artificial en el ámbito de la ciencia y partimos de la premisa de que la IA no produce conocimiento al margen de la sociedad, ya que aprende de datos generados por personas, se desarrolla bajo determinadas decisiones técnicas y empresariales y responde a valores, prioridades e intereses concretos de grandes desarrolladores tecnológicos. Por eso, puede amplificar sesgos, invisibilizar determinados conocimientos o reforzar narrativas erróneas. Si la IA va a convertirse en una nueva puerta de entrada de la información científica, deberíamos exigir transparencia sobre las fuentes que utiliza, los criterios con los que jerarquiza la información y los mecanismos con los que gestiona la incertidumbre y el error. La cuestión es quien decide cómo aprende, qué nos muestra y qué deja fuera la IA sobre conocimiento científico.