Two projects approved in FCT call to strengthen Public Administration through AI

Two projects led by VIS Team researchers, CertifyMeAP and BlockDFake, were approved under the FCT call for Artificial Intelligence, Data Science, and Cybersecurity of Relevance to Public Administration. Both initiatives, set to launch in March 2025, aim to address critical challenges in security and digital trust for public services.

CertifyMeAP, coordinated by Prof. Nuno Gonçalves, focus on combating forgery in professional training certificates issued by INA (National Institute of Administration), main responsible for providing training for Public Administration employees. Over 12 months, the project will digitize certifications by integrating AI, digital signatures, and printer-proof steganography. Collaborating INA, the team will develop a secure digital twin and mobile validation system to ensure reliable and accessible certification processes.

BlockDFake, led by PhD. Guilherme Schardong, targets the growing threat of illegal deepfake content. This 12-month project combines two strategies: detecting AI-generated video manipulations using implicit neural representations and embedding steganographic signatures for smartphone validation. Partnering with INCM (Portuguese Mint and Official Printing Office), the initiative will create a proof-of-concept architecture to protect video integrity through watermarking and real-time frame manipulation detection.