This research examines the critical balance of trust when integrating Large Language Models into software engineering workflows.

• Excessive trust can lead to security vulnerabilities and increased risks
• Insufficient trust may hinder innovation and adoption of beneficial AI tools
• Well-calibrated trust is essential as LLMs become integral to critical development processes
• Engineering impact: Provides a framework for responsibly implementing AI assistants in software development while managing potential risks

Mapping the Trust Terrain: LLMs in Software Engineering -- Insights and Perspectives