AI bot swarms threaten to undermine democracy:

Because humans update their views partly based on social evidence—looking to peers to see what is “normal”—fabricated swarms can make fringe views look like majority opinions. If swarms flood the web with duplicative, crawler-targeted content, they can execute “LLM grooming,” poisoning the training data that future AI models (and citizens) rely on. Even so-called “thinking” AI models are vulnerable to this.

We cannot ban our way out of the threat of generative-AI-fueled swarms of misinformation bots, but we can change the economics of manipulation.

The paper linked above explains the danger of the present moment where a single user or group can inexpensively use the existing social media platforms and LLMs to essentially poison democracies around the world.

They have some suggestions about what needs to change, but no specific steps.

Lots of folks have been worried about this for a long time. (Me too.) To my mind it’s an existential threat. If nothing else, it’s important to share this out so that people begin to be skeptical of what they seeing being posted online – especially when there appears to be coordinated outrage about something.