AI-Assisted Deception and the Emerging Challenge of LLMs in Forensic Psychiatry.

Abstract

Generative artificial intelligence (AI), including the large language model ChatGPT, has introduced potential new opportunities and challenges to the practice of forensic psychiatry. These powerful AI-based tools may offer substantial benefits in administrative tasks, report generation, and record summarization yet simultaneously present areas for further consideration, such as aiding evaluees in feigning psychiatric symptoms. Additional ethics and legal considerations exist regarding privacy, bias within AI models, and the introduction of fabricated or misleading AI-generated content into forensic assessments. Legislative efforts, privacy safeguards, and professional guidelines essential for responsible AI use are being developed. Forensic psychiatrists are uniquely positioned to influence responsible AI integration through education, advocacy, and development of best practices within psychiatry.