The project seeks to build professional trustworthy AI solutions against advanced disinformation techniques, co-created with and for media professionals and researchers and to also set the foundation for future research in the area of AI against disinformation. Key novel characteristics of the AI models will be fairness, transparency, robustness against concept drifts, continuous adaptation to disinformation evolution through a fact-checker-in-the-loop approach, and the ability to handle multimodal and multilingual content. Recognizing the perils of AI-generated content, the team will also develop tools for deepfake detection in all formats (audio, video, image, text).
ATC coordinates the development of vera.ai’s user-facing tools aiming to bring the AI-based research results to end-users from the media industry, social sciences, and other stakeholders like OSINT investigators. ATC is also responsible for the integration of state-of-the-art AI-based tools into Truly Media, through which these results will be made accessible to the European Digital Media Observatory (EDMO) and its communities. Furthermore, ATC will exploit its technical expertise to improve and enhance its social listening services, by conducting research and development in the area of identifying and tackling coordinated sharing behavior taking advantage of the latest advances in network science.
See more on the links between EDMO, the EDMO Hubs and vera.ai
The project follows on from previous projects projects where ATC’s Innovation Lab, the company’s dedicated R&D team has actively participated and worked, such as InVID (2016-2018), WeVerify (2018-2021) and Digger (2019 – 2020).
The vera.ai kick-off meeting was held in Greece, under the supervision of the Informatics Research Institute (ITI-CERTH), which is also the project’s coordinator.
You can find the original press release here.