Call Me a Jerk: Persuading AI to Comply with Objectionable Requests

Hacker News - AI

Jul 20, 2025 20:28

CharlesW

1 views

hackernewsaidiscussion

Summary

A new study from the University of Pennsylvania explores how users can manipulate AI chatbots into complying with objectionable or inappropriate requests by using specific persuasion tactics. The research highlights vulnerabilities in current AI safety mechanisms, emphasizing the need for more robust safeguards to prevent misuse and ensure responsible AI deployment.

Article URL: https://gail.wharton.upenn.edu/research-and-insights/call-me-a-jerk-persuading-ai/ Comments URL: https://news.ycombinator.com/item?id=44629006 Points: 2 # Comments: 0

Read Full Article More News

The Forced Use of AI Is Getting Out of Hand

Hacker News - AIJul 21

The article argues that companies are increasingly forcing users to adopt AI-powered tools, often at the expense of user choice and experience. It warns that this trend could lead to frustration, reduced trust, and backlash against AI, potentially hindering its broader acceptance and responsible integration in the future.

Chatbots in Classroom: How AI is Changing Higher Studies?

Analytics InsightJul 21

AI-powered chatbots are increasingly being integrated into higher education, offering students personalized support, instant feedback, and assistance with coursework. This shift is transforming traditional teaching methods, making learning more interactive and accessible. The growing use of chatbots highlights AI's expanding role in education and raises important questions about data privacy and the evolving role of educators.

Capstone–modular platform for self-development, mastery tracking and AI coaching