Alarmist title and poor article. I don’t agree that prompting an LLM to write targeted emails for your lobbying campaign counts as “weaponizing” AI —this is just common practice e.g. in sales and marketing.
If I disagree with you I can just pay to generate so much speech of mine that it will seem like nobody believes anything else. If that ability were used against you, I cannot imagine you not thinking that your opponent was wielding a powerful weapon against you