Developing deepfake literacy among pre-service teachers: a qualitative study in the context of immigration
This qualitative study demonstrates that while specific training in deepfake detection improves pre-service teachers' use of systematic verification strategies when evaluating immigration-related content, it does not fully eliminate reliance on unvalidated cognitive biases, underscoring the need for reinforced critical thinking in initial teacher education.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
In an age where a video can be made to look and sound exactly like a real person saying things they never said, the line between truth and fabrication is becoming dangerously thin. This phenomenon, known as a deepfake, relies on powerful computer programs that learn from vast amounts of real images and voices to create synthetic media. These tools can generate convincing videos of politicians, celebrities, or ordinary people doing or saying almost anything. While this technology offers new ways to create art and education, it also poses a serious threat to how we understand the world. When people cannot tell what is real, trust in news, institutions, and even each other begins to crumble. This is especially dangerous when it comes to sensitive topics like immigration, where false stories can inflame tensions and distort public debate. The challenge for society is not just to spot these fakes, but to teach people how to think critically about the information they consume every day.
A team of researchers in Spain set out to see if future teachers could learn to spot these digital forgeries, specifically in the context of immigration. They worked with 133 students training to become primary school teachers at the University of Murcia. These students were not experts in technology; they were ordinary people studying to guide the next generation. The researchers wanted to see how these students judged the truthfulness of digital content before and after receiving specific training on how to detect deepfakes. The goal was to understand what kind of reasoning people use when they try to decide if a video or image is real, and whether teaching them specific detection skills could change their approach.
Before the training began, the students were shown various pieces of digital content related to immigration and asked to decide if they were real or fake. At this stage, their methods were often unreliable and based on gut feelings rather than facts. Many students relied on what they already believed to be true, a mental shortcut known as confirmation bias. If a video about immigration fit their existing views or stereotypes, they were likely to accept it as real without checking further. Others trusted a piece of content simply because it looked official or came from a source they assumed was authoritative, such as a police chief or an old newspaper. Some even judged a video as fake just because the text looked poorly formatted or because the language seemed too emotional, without looking for actual evidence of manipulation. Their reasoning was often intuitive, meaning it felt right to them, but it was not grounded in a systematic check of the facts.
The researchers then introduced a training program designed to teach the students how to spot the specific signs of artificial intelligence manipulation. This instruction covered how to look for subtle errors that computers often make, such as a person's lips not moving in perfect sync with their voice, strange blinking patterns, or backgrounds that look blurry or distorted. The students also learned to check the source of the information, looking for verified news outlets rather than unverified social media accounts, and to cross-reference claims with reliable data. After this training, the students were shown the same types of content again and asked to judge them once more.
The results showed a clear shift in how the students approached the task. After the training, they moved away from relying on their initial feelings or stereotypes and began using more systematic methods. They were much more likely to look for technical glitches, such as the unnatural movement of a face or inconsistencies in the lighting. They also started checking the credibility of the source more carefully, asking whether the information came from a trusted news organization or a random social media post. While they still sometimes fell back on their own beliefs, the training gave them a new set of tools to question what they saw. The study suggests that specific education can help people become more careful and analytical when evaluating digital content, helping them to separate fact from fiction.
However, the researchers noted that the training did not completely eliminate the influence of personal beliefs. Even with new skills, some students still let their preconceptions guide their judgment. This highlights that while teaching people how to spot technical errors is helpful, it is not a complete solution. To truly protect against misinformation, especially on sensitive topics like migration, people need to combine technical skills with a strong habit of critical thinking. The study concludes that future teachers need this kind of training so they can pass these skills on to their own students, helping to build a society that is better equipped to navigate a world where seeing is no longer always believing.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.