by Cantivia
Artificial intelligence is growing increasingly powerful every day. These tools are cheap, fast, and highly sophisticated, and they are reshaping how disinformation and deepfakes spread. Our question was: could we fight against this prominent threat with our own set of AI-powered cybersecurity tools?
We are Cantivia, an Irish research group founded by three secondary-school students working on digital safety and the real-world impact of AI-generated disinformation. Rather than assume people can spot fakes, we wanted to conduct a survey to find out and to use what we learn to help build better defences.
To do that, we are running an anonymous, three-minute survey that is open to everyone. From the results, we are creating a synthetic population to simulate the virality of a piece of media. This is why we want to hear from as many respondents as we can. Every response directly improves our work and, by having a diverse representation in our survey, we can ensure that it is as accurate as possible.
Media literacy sits at the heart of this. Initiatives like Media Literacy Ireland’s Be Media Smart campaign have done vital work in encouraging people to stop, think, and check before trusting or sharing what they see. AI-generated content raises the stakes of that message, because the visual and audio cues we once relied on to catch a fake are disappearing. Understanding where people’s instincts still hold up, and where they break down, is essential if media literacy advice is going to keep pace with the technology.
For context, our work has been recognised with the Stripe Young Scientist 2026 Engineers Ireland STEPS Award, the Best Intermediate Technology Group Award, and a $20,000 Emergent Ventures grant to help take our research further.
If you have three minutes, we would be hugely grateful if you would take part, and share it with anyone who might also complete it. The survey is open until the end of August. Take the survey here.
Thank you for helping us understand, and push back against, the growing threat of AI-powered disinformation.