EmoDistill: Offline Emotion Skill Distillation for Language Model Agents in Adversarial Negotiation
arXiv:2605.26785v1 Announce Type: new Abstract: Post-trained LLMs are often optimized to align responses with human preferences, making them safe, polite, and conversationally appropriate. In adversarial negotiation, however, this alignment can become a vulnerability: emotionally framed language may steer agents toward the counterparty's interests. Using GoEmotions-based affective prompting, we show that emotion substantially shifts negotiation outcomes, suggesting that emotion is a strategic action channel rather than a surface style. Thus, we introduce \textbf{EmoDistill}, an offline framewo
The increasing sophistication of LLM applications, particularly in adversarial contexts, necessitates greater control over their strategic communication capabilities beyond mere safety alignment.
This research highlights that emotional expression in AI is not just a stylistic element but a strategic lever that can significantly influence outcomes in complex interactions like negotiation, impacting how AI agents can be directed or exploited.
The understanding that affective prompting and emotional skill distillation can be used to strategically manipulate adversarial outcomes, shifting LLMs from purely compliant systems to potentially opportunistic ones.
- · AI ethicists
- · Developers of strategic AI agents
- · Organisations performing automated negotiation
- · Entities interacting with sophisticated AI agents without emotional awareness
- · Developers of overly neutral or aligned LLMs
AI agents will become more adept at strategic influence through controlled emotional expression.
New cybersecurity vulnerabilities may emerge where AI agents exploit emotional responses or are themselves manipulated through targeted affective input.
The development of 'emotional AI intelligence' for both offense and defense in digital interactions could become a critical field, leading to an arms race in psychologically sophisticated AI.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.CL