Large language models can consistently generate high-quality content for election disinformation operations

Open in new window