UK Intelligence Community Postdoc award will help researcher detect AI-generated media and hidden messages

Dr Christian Schroeder de Witt is one of six engineering researchers in the UK awarded postdoctoral research grants that could help to advance national security. The latest UK Intelligence Community Postdoctoral Research Fellowships focus on unclassified basic research that can support the intelligence, security and defence communities.

The fellowships, which are offered by the Government Office for Science and administered by the Royal Academy of Engineering, provide a vital link between academia and the intelligence community. Each awardee receives funding for at least two years of their project and mentorship from a Fellow of the Academy as well as an advisor from the intelligence community.

Senior Research Associate Dr Christian Schroeder de Witt aims to identify AI-generated media, such as videos with audio, and detect hidden messages. 

As AI technology advances, it becomes easier to create fake or altered media (‘deepfakes’), which can be used for misleading or harmful purposes. Additionally, AI-generated media can be used to hide secret messages in ways that make them difficult to detect. The goal of the 'Detecting AI-generated media and hidden messages therein' project is to develop new methods to detect both AI-generated content and any hidden information within it. By creating more effective tools, Dr Schroeder de Witt and his team can better protect against the misuse of AI in spreading false information or concealing secret communications. This research will also provide resources and tools to the broader community to help improve the detection and understanding of AI-generated media.

Dr. Schroeder de Witt recently appeared live in-studio on BBC News to discuss AI's role in combating disinformation - read more here.

Read the RAEng announcement