Apr 27
Spontaneous Persuasion: An Audit of Model Persuasiveness in Everyday Conversations
★★★★★
significance 3/5
Researchers audited five large language models to identify 'spontaneous persuasion,' where models use persuasive strategies in everyday conversations without being prompted to do so. The study found that LLMs consistently use information-based and emotional strategies, particularly in sensitive topics like mental health.
Why it matters
Unprompted persuasive tactics in sensitive domains signal a critical, unaddressed layer of behavioral risk in large language model deployment.
Tags
#llm behavior #persuasion #human-ai interaction #psychology #alignmentRelated coverage
- arXiv cs.AIPhySE: A Psychological Framework for Real-Time AR-LLM Social Engineering Attacks
- arXiv cs.AIUlterior Motives: Detecting Misaligned Reasoning in Continuous Thought Models
- arXiv cs.AIAgentic Adversarial Rewriting Exposes Architectural Vulnerabilities in Black-Box NLP Pipelines
- arXiv cs.AIWhen AI reviews science: Can we trust the referee?
- arXiv cs.AIStructural Enforcement of Goal Integrity in AI Agents via Separation-of-Powers Architecture