‘Inoculation’ helps people spot political deepfakes, study finds

Source Domain: www.kpvi.com

The study highlights the effectiveness of both text-based warning messages and interactive games in improving people’s ability to spot AI-generated political deepfakes.
Inoculation theory suggests that informing people about how deepfakes work can help them recognize and resist these misleading AI-generated videos and audio.
The experiment had three groups: one receiving passive text warnings, another an active interactive game challenging them to identify deepfakes, and the third receiving no inoculation.
The research found that both inoculation methods reduced participants’ credibility of deepfakes and increased awareness and intention to learn more about detecting them.
Deepfakes pose a serious risk to democracy by creating realistic fake videos and audio that politicians could appear to say, potentially harming public trust and spreading false information.
While fact-checking deepfakes is currently ineffective due to the rapid spread of misinformation, inoculation offers a promising alternative by preparing people to resist such AI-generated content.
The study contrasts with previous research that focused mainly on text-based media literacy approaches and mainly text-based misinformation.
Future studies aim to explore the long-term effectiveness of inoculation messages and its applicability beyond political misinformation, such as in health contexts.

‘Inoculation’ helps people spot political deepfakes, study finds | Interests