AI significantly amplifies cybersecurity threats. It can already be used to conduct convincing phishing campaigns, impersonate trusted co-workers with voice cloning, and perform large-scale analysis of open source code for vulnerabilities. In the future, these attacks will become cheaper, faster, and more accessible for cybercriminals, and each new generation of AI models will enable additional threats.
The CivAI demos aim to provide a tangible understanding of just how easy AI will make things for cybercriminals. Deep, practical awareness of these threats can better inform cybersecurity policy as it strives to keep pace with recent AI progress. To prevent any possibility of misuse, the demos are not public, but approved organizations can view them by contacting CivAI using the form above.
See a past edition of our newsletter for a real-world example of an AI-powered scam in which a fabricated video was created to exploit the popularity of MrBeast, the highest-paid celebrity on YouTube.
Stay tuned for a public announcement of our next demo, This Is Not You. This web application will allow anyone to make a deepfake of themselves while learning about the impacts of AI technology.
