Observational analysis of a large dataset of AI agent posts on Moltbook identifies 18.28% harmful content and 74 malicious behavior classes alongside evidence of coordinated posting campaigns.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
When Agents Talk: Discourse, Manipulation, and Risk in an Agentic Social Network
Observational analysis of a large dataset of AI agent posts on Moltbook identifies 18.28% harmful content and 74 malicious behavior classes alongside evidence of coordinated posting campaigns.