Combating disinformation in the age of generative AI : from watermarking LLMs to persuasion analysis of memes